Home

RacerPro User's Guide Version 1.9.2

1. constrained peter age of peter age gt OKAY constraints age of peter 36 Vv OKAY retrieve x told value if exists age x x and gt age 30 lt age 40 x peter existing told values age x 36 Vv Racer Systems GmbH amp Co KG http www racer systems com 112 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL Note that age of peter is called a concrete domain object CD object and 36 is a concrete domain value Please note that concrete domain are in fact variables in a concrete domain constraint network nRQL tries to identify and maximize as many told values as possible constrained mary age of mary age gt OKAY N constraints age of peter age of mary OKAY v retrieve x told value if exists age x x and gt age 30 lt age 40 gt x mary existing told values age x 36 x peter existing told values age x 36 However in some case this is not possible in this case told value if exists returns the CD object instead of the told value constrained paul age of paul age gt OKAY N constraints lt age of paul age of mary OKAY v 9 retrieve x told value if exists age x x lt age 40 gt x peter existing told values age x 36 x paul existing told values age x age of paul x mary existing told values age x 36
2. full reset okay full reset Vv implies mother woman OKAY Vv define concept mother woman with children Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 161 gt OKAY atomic concept synonyms mother gt woman with children mother N taxonomy top NIL woman woman with children mother woman bottom woman top mother woman with children bottom mother woman with children NIL v We construct a special relational structure from the taxonomy which is then queried with nRQL see right hand side of Figure 6 1 This relational structure can also be seen as virtual taxonomy ABox Virtual means that this is not a real ABox However under this perspective TBox queries become easy to understand So a TBox query is just an ordinary ABox query which is evaluated on the virtual taxonomy ABox The taxonomy ABox mirrors the taxonomy graph see Figure 6 1 In the taxonomy there is one node representing a whole set of synonym concepts e g mother and woman with children is represented as one node One of these synonym concepts is called the representative concept The representative concepts are shown as bold non dashed circles in Figure 6 1 Thus the taxonomy ABox contains one individual node for each concept name The synonym individuals are then linked by means of has synonym edges role assertions Moreover the individual
3. gt inconsistent describe query last gt query 1 accurate processed retrieve x and x woman x man abox smith family reexecute query last gt inconsistent prepare abox query Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 221 x and x woman x y has child y all has parent all has descendant woman y z has child z u has child u uncle NRQL WARNING query 2 IS INCONSISTENT gt query 2 ready to run prepare abox query x and x woman x y has child y all descendant of all has descendant woman y z has child z u has child u uncle gt query 3 ready to run N define primitive role descendant of inverse has descendant gt OKAY prepare abox query x and x woman x y has child y all descendant of all has descendant woman y z has child z u has child u uncle gt query 4 ready to run N execute query last gt inconsistent retrieve x union and x man x woman and x grandmother x uncle NRQL WARNING query 5 IS INCONSISTENT gt inconsistent Racer Systems GmbH amp Co KG http www racer systems com 222 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL retrieve nil and x grandmother neg x woman NRQL WARNING query 6 IS INCONSIS
4. instance x mother was fired in advance So let s add these children with a rule firerule and x mother neg x has known successor has child instance new ind first child of x human related x new ind first child of x has child gt instance first child of eve human related eve first child of eve has child instance first child of doris human related doris first child of doris has child Note that the operator new ind lt name gt lt ind or var gt creates a new ABox individ ual prefixed with lt name gt for each different binding possibility of the argument objects in lt ind or var gt and new ind lt name gt will simply create a new ABox individual lt name gt Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 155 Note that a rule with an empty antecedence nil or is conceptually a conjunction without arguments and thus interpreted as TRUE thus also true query could be used A rule with tautological always true precondition can be used to add assertions But please note that the objects mentioned in the consequence must also appear in the precondition thus for firerule instance new ind person you get NRQL ERROR Parser Error Object new ind not mentioned in query body nil Please use firerule instance new ind new ind person instead 6 1 6 5 Rules and the Concrete Domain Rules can al
5. lt tbox query head gt gt lt tbox head entry gt Racer Systems GmbH amp Co KG http www racer systems com 194 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL lt abox head entry gt lt tbox head entry gt lt def head entry gt lt query object gt lt lambda application gt lt lambda var gt lt lambda body gt lt lambda argument gt lt query variable gt lt query individual gt lt data query object gt lt data query variable gt lt query object gt lt data query object gt lt abox head projection operator gt lt lambda application gt lt query object gt lt tbox head projection operator gt lt lambda application gt lt query object gt lt data query object gt lt query variable gt lt query individual gt lambda lt lambda var gt lt lambda body gt lt lambda argument gt lt symbol gt A MiniLisp S expression see the Manual lt query object gt lt data query object gt lt head projection operator gt lt symbol gt lt symbol gt lt symbol gt naming an ABox individual or concept lt data query variable gt lt data query individual gt gt k lt symbol gt lt symbol gt lt data query individual gt gt lt symbol gt Data query objects can only be used if nRQL is used in hybrid mode i e a data substrate mode is enabled Racer Systems GmbH amp Co KG h
6. x alice v retrieve x meg same as x alice x doris x charles x betty x eve v Note that same as works on a syntactic level unless you turn on use individual synonym equivalence classes see also the output of describe query processing mode A semantic equality predicate is given by role query atoms using the nrql equal role 6 1 4 3 6 Negation as Failure for Binary Atoms with Individuals Both role query atoms same as query atoms and constraint query atoms are binary atoms In principle a binary atom and its NAF negated variant are complementary However since these atoms reference two objects the effect of the final projection must be taken into account Moreover full reset okay full reset v N racer read file OKAY nrql user guide examples family racer Vv retrieve alice x alice x has child alice alice x betty alice alice x charles v N retrieve alice x neg alice x has child v alice eve x eve alice eve x doris alice eve x charles alice eve x betty alice eve x alice alice doris x eve alice doris x doris alice doris x charles alice doris x betty alice doris x alice alice charles x eve alice charles x doris alice charles x charles alice charles x betty alice charles x alice
7. x doris y charles x doris y betty x doris y alice x charles y eve x charles y doris x charles y betty x charles y alice x betty y eve x betty y doris x betty y charles x betty y alice x alice y eve x alice y doris x alice y charles x alice y betty retrieve x y x y constraint age age lt gt gt x alice y doris x alice y eve x alice y betty x alice y charles x doris y alice x doris y eve x doris y betty x doris y charles x eve y alice x eve y doris x eve y betty x eve y charles x betty y alice x betty y doris x betty y eve x betty y charles x charles y alice x charles y doris x charles y eve x charles y betty Thus both queries return the same answer However the situation changes is also role chains are used within constraint query atoms in this case the complement can no longer be retrieved by simply negating the CD predicate since additional implicit constraints resp retrieval conditions on the bindings are imposed by the role chains Again the semantics of neg is defined in such a way that a constraint query atom and its negated variant are always complementary Racer Systems GmbH amp Co KG http www racer s
8. Cimplies top all has Cimplies some has chile Cimplies some has sibli Cimplies top all has 4 Cimplies top Call has sister some has gender female Cimplies top all has brother some has gender male Cimplies person and human some has gender or female male disjoint female male CimpLies woman and person some has gender female Cimplies man Cand person some has gender male equivalent parent and person some has child person equivalent mother Cand woman parent equivalent father and man parent equivalent grandmother Cand mother some has child some has child person equivalent aunt and woman some has sibling parent equivalent uncle and man some has sibling parent v Select Buffer family owl localhost 8088 family racer Lisp Users rm Desktop family racer 4 Figure 2 3 RacerEditor displaying the list of buffers RacerPro supports the standard Emacs commands such as for instance change buffer etc In addition the standard key bindings are provided E g change buffer is bound to c x b Press tab to see a list of possible completions see Figure 2 3 Racer Systems GmbH amp Co KG http www racer systems com 2 1 SAMPLE SESSION WITH RACERPORTER 15 After having loaded the file family racer into RacerPro you can use RacerPorter to inspect the knowledge base For instance you might be interested in the hierarchy of concept names or
9. Figure 2 13 Second step for declaring RacerPro as the reasoner used by Prot g family 1 Prot g 3 1 beta file Users rm Desktop family 1 pprj OWL Files owl or rdf OWL Preferences _ Disable Drag and Drop Reasoning Reasoner URL http localhost 8080 General Visibility Searching Encoding User Interface Features Icon Style Default Protege Features __ Import Protege metadata ontology Language Profile Use standard profile OWL Full _ O Use custom profile AETI In the preferences dialog window specify host and port number see Figure 2 13 Usually the defaults localhost 8080 are ok Another example would be 135 28 70 104 8081 After making required changes press the close button Racer Systems GmbH amp Co KG http www racer systems com 34 CHAPTER 2 USING RACERPRO le 00 family 1 Prot g 3 1 beta file Users rm Desktop family 1 pprj OWL Files owl or rdf File Edit Project OWL Code Window Tools Help n J Deb Efh e Aea Se E BB lt K prot g OWLClasses W Properties Forms Individuals Metadata SUBCLASS RELATIONSHIP gt mA CLASS EDITOR Poe Tt For Project farnily 1 For Class family UNCLE instance of owl Class asserted Hierarchy 8 P A A Name Sameas DiterentErom J annotations Cem owl Thing family UNCLE lel Property Value Lang family FEMALE
10. Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 227 6 2 8 The Query Repository The QBox The query entailment check see previous Section is used for the maintenance of a query repository This repository is also called the QBoz If enabled a QBox is maintained for each queried ABox The QBox is created on demand if needed The QBox serves as a hierarchical cache The service is still experimental and should be considered an non essential add on For each new query to be answered its set of most specific subsumers as well as its most general subsumes are computed This process is called query classification The QBox can be seen as a taxonomy for queries If a classified query is about to be executed the cached answer sets of its parent queries direct subsuming queries can be utilized as superset caches and the cached answer sets of its children queries direct subsumed queries serve as subset caches In case the QBox already contains an equivalent query with a cached answer set this set will be returned immediately The query entailment check which is used for query classification is slightly stricter than the query entailment check just discussed Here we require in addition that the object vectors must have the same length The rationale is that we want to exploit and directly reuse the cached tuples the tuples from the caches must therefore have the same
11. family HUMAN rdfs comment P family MALE v family MAN family BROTHER family FATHER family UNCLE v familyPERSON v family PARENT Asserted Inferred im Properties ew wu e a family FATHER y R eal Asserted Conditions Ci ud vi Peel 7 om fa MAN family GRANDMOTHER NECESSARY amp SUFFICIEN Y family WOMAN family MAN famil AUNT 3 family HAS SIBLING family PARENT Y family MOTHER rame ORANDMOTHER 3 family HAS GENDER family MALE m family A farnily SISTER D Disjoints OE E 3 I i atida Logic View Properties View Figure 2 14 Prot g displaying the Family knowledge base The class UNCLE is selected and in the subwindow for asserted conditions necessary and sufficient conditions are shown graphically The example presented in Figure 2 14 shows a screenshot of Prot g with the Family knowl edge base The knowledge base was exported from RacerPro using the OWL syntax see the function save kb OWL files can be manipulated with Prot g In Figure 2 14 the concept uncle is selected concepts are called classes in Prot g See the restrictions displayed in the asserted conditions window and compare the specification with the KRSS syntax used above see the axiom for uncle Note that the class window displays only obvious subclass relationships As we will see later RacerPro can be used to also c
12. grandpa and man some has child parent great grandpa and man some has child some has child parent Racer Systems GmbH amp Co KG http www racer systems com 244 APPENDIX B A KNOWLEDGE BASE WITH CONCRETE DOMAINS equivalent grandma and woman some has child parent equivalent great grandma and woman some has child some has child parent equivalent aunt and woman some has sibling parent equivalent uncle and man some has sibling parent equivalent sibling or sister brother equivalent mother with siblings and mother all has child sibling equivalent brother and man at least 1 has sibling equivalent sister and woman at least 1 has sibling implies at least 2 has child all has child sibling implies some has child sibling at least 2 has child implies sibling all inv has child and all has child sibling at least 2 has child equivalent mother having only sisters and mother all has child and sister all has sibling sister Alice is the mother of Betty and Charles instance alice and woman at most 2 has child Alice s age is 45 constrained alice age of alice has age constraints equal age of alice 45 related alice betty has child related alice charles has child Betty is mother of Doris and Eve instance betty and woman at most 2 has child Betty s age is 20 constrained betty age of betty has a
13. let w reduce flatten execute or reexecute query weights of parts of car Racer Systems GmbH amp Co KG http www racer systems com 124 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL parts length execute or reexecute query parts of car car car total weight w car gt car cari no of parts 4 total weight 630 0 6 1 3 3 3 User Defined Query Result Format As already demonstrated the value returned by a lambda body is included in the binding list In the previous query the lambda body returns a structured list using the template car car total weight w this is equivalent to list list car car list total weight w It is also easy to specify natural language output simply replace car car total weight w with format nil Car with name A weights A kg and has A parts car w parts in the previous query and you will get Car with name cari weights 630 0d0 kg and has 4 parts as the query result 6 1 3 3 4 File Output MiniLisp also offers the with open output file which allows to open an output file Simply print to the stream output stream which is established by with open output file to add arbitrary content to the output file retrievet car car lambda car with open output file minilisp output txt let w reduce flatten retrievet and car car car part has part part a w
14. lt owl FunctionalProperty rdf ID name gt lt rdfs range rdf resource http www w3 org 2001 XMLSchema string gt lt rdfs domain rdf resource person gt lt rdf type rdf resource http www w3 org 2002 07 owl DatatypeProperty gt lt owl FunctionalProperty gt lt person rdf ID alice gt lt age rdf datatype http www w3 org 2001 XMLSchema int gt 80 lt age gt lt name rdf datatype http www w3 0rg 2001 XMLSchema string gt Alice lt name gt lt hasChild gt lt person rdf ID betty gt lt age rdf datatype http www w3 0rg 2001 XMLSchema int gt 50 lt age gt lt name rdf datatype http www w3 0rg 2001 XMLSchema string gt Betty lt name gt lt person gt lt hasChild gt lt person gt lt rdf RDF gt First we like to mention that it is possible to query datatype properties as if they were concrete domain attributes we have extended the RacerPro concept expression syntax ap propriately retrieve x x gt age 75 gt x http www owl ontologies com unnamed owl alice retrieve x x string name Betty Racer Systems GmbH amp Co KG http www racer systems com 106 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL gt x http www owl ontologies com unnamed owl betty retrieve x x a name 35 gt x http www owl ontologies com unnamed owl alice x http www owl ontologies com unnamed owl betty Constraint query a
15. x charles As the name suggests union simply returns the union of the query answers returned by its argument bodies disjuncts A disjunct which references n objects variables and or individuals denotes a set of n ary tuples Since the union operator is only meaningful if all disjuncts have the same arity nRQL nRQL ensures that the disjuncts will always have the same arity This is achieved by adding additional lt variable gt top conjuncts to disjuncts retrieve x y union x y has child x mother gt x alice y eve x alice y doris x alice y charles x alice y betty x alice y alice x betty y eve x betty y doris x betty y charles x betty y betty x betty y alice describe query last gt query 10 accurate processed retrieve x y or and x mother top y x y has child Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 129 abox smith family Note that nRQL has rewritten the original query body into or and x mother top y x y has child It is important to note that the variable names matter Consider the following query retrieve x union x man y woman gt x charles x eve x doris x betty x alice A common pitfall is to think that this query is equivalent to the query retrieve y y man
16. PN NNO node name http www owl ontologies com Ontology1159352693 owl test node label top http www owl ontologies com Ontology1159352693 0wl class abox object abox individual owl object owl individual node successors 123 Comment for individual test 456 http www owl ontologies com Ontology1159352693 0wl test node predecessors http www owl ontologies com Ontology1159352693 owl test describe all edges gt from node http www owl ontologies com Ontology1159352693 0wl test to node 123 edge label owl relationship owl datatype property relationship http www owl ontologies com Ontology1159352693 owl datatype property from node http www owl ontologies com Ontology1159352693 0wl test to node Racer Systems GmbH amp Co KG http www racer systems com 184 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL http www owl ontologies com Ontology1159352693 owl test edge label owl object annotation property relationship http www owl ontologies com Ontology1159352693 owl annotation object property abox relationship abox role relationship owl relationship owl object property relationship http www owl ontologies com Ontology1159352693 owl object property from node http www owl ontologies com Ontology1159352693 0wl class to node This is a test class edge label owl relationship owl data
17. abox attribute relationship age get data edge description alice age 80 gt from node alice age to node 80 edge label abox relationship abox told value relationship get data edge description alice name Alice gt from node alice name to node Alice edge label abox relationship abox told value relationship Please consider the edge label information The meaning of the markers should be obvi ous given the previous explanations for the node labels Again there are three types of edges 1 Edges connecting nodes representing ABox individuals e g alice and betty such an edge is created due to a role membership ABox assertion 2 Edges connecting a node representing an ABox individual with a node representing a concrete domain CD object which is an attribute filler of the ABox individual e g alice and alice name Such an edge is caused by a constrained ABox assertion Racer Systems GmbH amp Co KG http www racer systems com 178 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 3 Edges connecting a node representing a CD object with node representing a told CD datatype value e g alice age and 80 Such an edge is caused by a constraints assertion Please also try describe all nodes describe all edges We can now exploit the more expressive representation using hybrid nRQL The same data query atoms as already introduced in the previous secti
18. for RDF OWL files Lisp for RacerPro files and Fundamental for all others as well as the path of the imported file Please notice that the Editor is similar to Emacs and therefore not behaving like typical Windows based text processor For instance browsing through the text with the help of the window scroll bar may move the cursor too Also you can not select the text by holding down the shift key and move the cursor with the arrow keys Rather highlight the text by clicking and shift dragging the mouse over the interesting passages then right click the mouse to select if you want to copy or cut the text from the window Remember that the control key especially the control C keystroke combination has a different meaning in Emacs and analogously in RacerEditor In the following we list the most important key bindings Racer Systems GmbH amp Co KG http www racer systems com 28 CHAPTER 2 USING RACERPRO e Return or Enter fresh line or send command e Tab Completion key e Ctrl a Beginning of line e Ctrl e End of line e Ctrl k Kill line e Ctrl left Matching starting e Ctrl right Matching ending e Ctrl d Del Delete e Backspace Backspace e Ctrl Space Set Mark e Meta w Copy e Ctrl y Paste e Ctrl w Cut e Meta p Alt p Previous Command e Meta n Alt n Next Command 2 4 3 Tabs in RacerPorter In the Taxonomy tab select a concept from the taxonomy then use
19. lt rdf first gt lt swrl IndividualPropertyAtom gt lt swrl propertyPredicate rdf resource hasSibling gt lt swrl argument2 rdf resource z gt lt swrl argumenti rdf resource y gt lt swrl IndividualPropertyAtom gt lt rdf first gt lt rdf rest rdf resource http www w3 org 1999 02 22 rdf syntax ns nil gt lt swrl AtomList gt lt swrl head gt lt swrl Imp gt lt Person rdf ID alice gt lt hasChild gt lt Person rdf ID betty gt lt hasChild gt Racer Systems GmbH amp Co KG http www racer systems com 249 lt hasChild gt lt Person rdf ID charles gt lt hasChild gt lt Person gt lt rdf RDF gt Racer Systems GmbH amp Co KG http www racer systems com 250 APPENDIX C SWRL EXAMPLE ONTOLOGY Racer Systems GmbH amp Co KG http www racer systems com Appendix D LUBM benchmark 333 Mode Lisp Syntax Ansi Common Lisp Base 10 in package CL USER setf logical pathname translations racer lubm Cc Ralf LUBM edit this line defconstant no of runs per query 1 defmacro benchmark universities departments no query vars let t1 get internal real time number of answers nil id intern format nil LUBM QUERY A no dotimes i no of runs per query racer prepare query vars query id id setf number of answers get answer size id t let time
20. Then evaluate compile load lracer Lracer to compile and load LRacer You have to use the pathname that corresponds to your LRacer distribution directory of course Afterwards you can import the package RACER into your own package or access RacerPro directly from the Common Lisp User package and the Lisp listener The variable default racer host can be set to a string denoting the host on which the RacerPro server run the default is localhost You can set the tcp port for the RacerPro server by setting the variable default racer tcp port default value is 8080 You can explicitly open a server connection with the function open server connection In order to close the server connection use the function close server connection See also the macro with server connection However explicitly opening a server is not necessary it just reduces the network overhead for multiple server calls 2 2 3 Web Service Interface A web service interface is provided with the OWL QL server The documentation for this software can be found at http www racer systems com 2 2 4 HTTP Interface DIG Interface The DIG Interface is a standardized XML interface to description logics systems developed by the DL Implementation Group DIG see http dig sourceforge net for details In a similar way as the socket interface the HTTP interface can be used from application programs and graphical interfaces If the option f is not provided the HTTP
21. and cap intersection lt abox query body gt or cup union lt abox query body gt not neg lt abox query body gt inv lt abox query body gt lt tbox query body gt gt lt empty query body gt Racer Systems GmbH amp Co KG http www racer systems com 196 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL lt tbox query atom gt wen project to project pi lt def query head gt lt tbox query body gt and cap intersection lt tbox query body gt or cup union lt tbox query body gt not neg lt tbox query body gt iny lt tbox query body gt lt empty query body gt gt lt abox query atom gt gt true query false query not neg lt abox query atom gt lt abox query atom gt inv lt query object gt lt concept expression gt l lt query object gt lt query object gt lt role expression gt nrql equal role lt query object gt nil lt role expression gt nil lt query object gt lt role expression gt l top lt query object gt bottom lt query object gt lt query object gt lt query object gt constraint lt role chain followed by attribute gt lt role chain followed by attribute gt lt predicate expressio
22. eee TopBraid pizza owl Eclipse SDK Users rm Documents workspace rie 12 aA n Je gt Cheeseyrizea ay EY TopBraid amp Java Te Classes INS se OC D ff properties 22 o v rafs Resource y g S njaa fg mi mv ow AlDifferent dla S S MLA fo 4 I hasCountryOfOrigin A v canis gt I hasingredient o CheeseyPizza OM hasSpiciness e Faea gt I isingredientof IM protege FROM s pn lM protege SLOT CONSTRAINTI BR protege TO NonvegetarianPizza E protege PAL DESCRIPTION owl Nothing rdfs Resource E protege PAL NAME protege DIRECTED BINARY RELATIO E protege PAL RANGE H protege PAL CONSTRAINT 7afesubClassot MM protege PAL STATEMENT H panna ows Thing E owiversioninfo picyPizza WR protege abstract gt SpicyPizzaEquivalent owkunionOF W protege allowedParent SpicyTopping gt Laer am WR protege defaultLanguage dirype MM protege excludedTest E protege probeClass MO protege readOnly MO protege subclassesDisjoint MR protege todoPrefix E protege todoProperty MM protege usedLanguage gt valuePartition VegetarianPizza VegetarianPizzaEquivalent1 VegetarianPizzaEquivalent2 VegetarianTopping gt rdfProperty raf statement hasTopping some CheeseTopping and Pizza CheeseyPizza ecuivalentClace gt rafs Class owitclass E rdfs comment I rdfs label gt E rdfs seeAlso E x
23. get internal real time t1 internal time units per second no of runs per query format t Universities 2 D Max Deps 2 D Query 2 A Answers 7 D Time A 1 universities if null departments all 1 departments no if null number of answers 0 number of answers float time defun load kbs n universities max n departments let kb name owl read file namestring translate logical pathname racer lubm university univ bench 0wl verbose t kb name lubm Racer Systems GmbH amp Co KG http www racer systems com 252 APPENDIX D LUBM BENCHMARK loop for i from 0 to n universities as n departments 1 length directory namestring translate logical pathname format nil racer lubm university university A owl i do loop for j from 0 to if null max n departments n departments min max n departments n departments as filename namestring translate logical pathname format nil racer lubm university university A A owl i j do owl read file filename verbose t init nil kb name kb name ignore import t defun prepare lubm data n universities check abox consistency n universities max n departments let tia get internal real time time load kbs n universities max n departments let tib get internal real time let t2a get internal real time format t Z ZABox preparation time prepare abox fo
24. gt OKAY implies grandmother mother gt OKAY define concept mother and woman some has child top gt OKAY instance alice grandmother gt OKAY instance betty or mother grandmother gt OKAY related betty eve has child gt OKAY instance betty all has child woman gt OKAY Let us switch to mode 4 and retrieve the instances of woman set nrql mode 4 gt okay mode 4 retrieve x x woman gt query 3 running get next tuple last gt x alice get next tuple last gt warning expensive phase two starts get next tuple last gt x betty get next tuple last gt x eve Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 217 Now let us switch to mode 5 Please observe that phase one is no able to produce one more tuple before warning expensive phase two starts is delivered set nrql mode 5 gt okay mode 5 retrieve x x woman gt query 4 running get next tuple last gt x alice get next tuple last gt x betty get next tuple last gt warning expensive phase two starts get next tuple last x eve Vow 6 2 5 4 Mode 6 vs Mode 3 Mode 6 is like mode 3 in set at a time mode However unlike mode 3 in mode 6 nRQL will exploit the two phase query processing schema Thus the more expens
25. owl Thing v lo e 8 amp 8 Logic View Properties View 4 Figure 2 34 Necessary and sufficient conditions for French or italian e 00 french italian Prot g 3 1 1 file Applications Protege_3 1 french italian pprj OWL Files owl or rdf N bee EEERrPpEEEDIN EK prot g Properties Forms lt Individuals amp Metadata S SWRL Rules SUBCLASS RELATIONSHIP CLASS EDITOR For Project french italian For Class No wine drinker instance of owl Class Asserted Hierarchy ds 18 Name SameAs DifferentFrom LL Annotations owl Thin Property y ka No wine drinker 9 swrkimp swrl Variable rdfs comment P gt swrl Atom P swrl Builtin French Italian a French or Italian No wine drinker Asserted Conditions gE pd NECESSARY amp SUFFICIENT gt Wine drinker owl Thing a O 7 aJs elek Logic View Properties View 77 Figure 2 35 Necessary and sufficient conditions for Not wine drinker Racer Systems GmbH amp Co KG http www racer systems com 52 CHAPTER 2 USING RACERPRO 0909 french italian Prot g 3 1 1 file Applications Protege_3 1 french italian pprj OWL Files owl or rdf Dl e e elsia gt oo Sl lt gt i prot g OWLClasses P Properties Forms lt D gt Individuals amp Metadata SWRL Rules CLASS BROWSER INST
26. retrieve x x mother gt x alice x betty x doris x eve Racer Systems GmbH amp Co KG http www racer systems com 154 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL Note that rules using NAF negated atoms like neg x mother behave non monotonically Because now there are no more woman which are not also mothers the rule can t be applied a second time firerule and x woman neg x mother instance x mother gt NIL 6 1 6 3 Controlling what to add to an ABox RacerPro can be advised to automatically add created rule consequences to an ABox if add rule consequences automatically is enabled you must add rule consequences manually see below if dont add rule consequences automatically has been evalu ated use describe query processing mode to learn which mode is currently enabled Moreover the rule API allows a fine grained control over what and when something is added 6 1 6 4 Creating New ABox Individuals with Rules A rule can also create new individuals in an ABox For example there are individuals in the ABox which are mothers but there are no known children modeled retrieve x and x mother neg x has known successor has child gt x doris x eve Thus doris and eve are known to be mothers but they have no known children If you are getting NIL please ensure that the rule firerule and x woman neg x mother
27. x doris existing told values age x 24 x eve existing told values age x 18 Ko Please note that age of paul is returned since the told value could not be identified This is not surprising since Paul s age is underspecified we only know that he is younger than Mary There are two more operators which address concrete domain objects Operators of the form lt attribute gt lt object gt e g age x return the concrete domain object which is the filler of lt attribute gt of the current binding of lt object gt retrieve x age x x an age gt x peter age x age of peter x paul age x age of paul x mary age x age of mary x alice age x no cd objects x doris age x no cd objects Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 113 x eve age x no cd objects x betty age x no cd objects x charles age x no cd objects The told value operator can be applied to CD objects if existing to retrieve told values for that CD object retrieve x told value age x x an age gt x peter told values age x 36 x paul told values age x no told value x mary told values age x 36 x alice told values age x no cd objects x doris told values age x no cd objects x eve tol
28. x http www owl ontologies com unnamed owl boo0k123 Please note that it is important to use the correct concept names the name http www owl ontologies com unnamed owl person is a so called qualified name and http www owl ontologies com unnamed owl is the so called XML namespace perfix The namespace prefix can be retrieved with the function get namespace prefix In order to avoid typing of these long names the prefix may be used to abreviate the current namespace prefix see the last query Note that this does not only apply to concept class names but also to individuals and roles properties retrieve book123 book gt t If you are using an older version lt 1 9 0 of RacerPro RacerPorter you must use full qualified names for the classes properties and individuals doesn t work and also put bars around names like in this example retrieve x x http www owl ontologies com unnamed owl person gt x http www owl ontologies com unnamed owl michael1 In case you are unsure about the correct names use the following functions to find out all atomic concepts gt top bottom http www owl ontologies com unnamed owl person http www owl ontologies com unnamed owl book all roles gt inv http www w3 org 2000 01 rdf schema comment Racer Systems GmbH amp Co KG http www racer systems com 98 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE
29. Co KG http www racer systems com 174 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL retrieve x y x y satisfies predicate lt gt x betty y alice Here we have been searching for substrate nodes x y which satisfy the binary predicate lt this means that their node labels must contain numeric literals a and b such that a lt b holds A satisfies predicate must always have arity two Please refer to 193 rule data edge query satisfies expression for a list of currently supported predicates So the query is somehow the equivalent of this concrete domain query retrieve x y x y constraint age age lt gt x betty y alice 6 1 8 2 The Mirror Data Substrate Most users will prefer a facility which automatically creates associated data substrate objects for ABox individuals This is what the mirror data substrate does The mirror data substrate is also very useful for users who want to query OWL KBs with nRQL If the facility is enabled RacerPro will automatically create data substrate objects for all elements in an OWL KB 6 1 8 2 1 Populating and Querying the Mirror Data Substrate Let us consider an example to demonstrate the benefits of the mirror data substrate for this purpose we must augment family no signature racer a little bit We also need some constrained as well as constraint ABox assertions 7 full reset okay full r
30. In this case some applications require that this might be indicated by a certain event For instance in a document retrieval scenario an application submitting an instance retrieval query for searching documents might also state that future matches should be indicated In order to support these features RacerPro provides the publish subscribe facility The idea of the publish subscribe system is to let users subscribe an instance retrieval query under a certain name the subscription name A subscribed query is answered as usual i e it is treated as an instance retrieval query The elements in the result set are by definition are only those individuals of the A box in question that have been published previously If information about a new individuals is added to an A box and these individuals are published the set of subscription queries is examined If there are new elements in the result set of previous queries the publish function returns pairs of corresponding subscription and individual names 5 3 1 An Application Example The idea is illustrated in the following example taken from a document retrieval scenario In some of the examples presented below the result returned by RacerPro is indicated and discussed If the result of a statement is not discussed then it is irrelevant for understanding the main ideas of the publish subscribe mechanism First a T box document ontology is declared in tbox document
31. Moreover output stream is bound to the file output stream within the scope of with open output file this is an ordinary CL output stream and can thus be used as argument to standard CL functions such as format See below for examples for file output The following special forms are provided reduce and or not if when unless cond maptree maplist every some progn progi let let lambda with nrql settings Note that functions like maptree take lambda bodies as argu ments however lambda expressions are not first order in MiniLisp in order to grant termination These head projection operators are available as functions describe ind instantiators most specific instantiators retrieve individual synonyms Some type conversion functions to number to string to symbol There are some predefined sorting functions sort string greaterp sort string lessp sort string lt sort string gt sort symbol name greaterp sort symbol name lessp sort lt sort gt Racer Systems GmbH amp Co KG http www racer systems com 120 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL The following functions are borrowed from Common Lisp and work as expected 1 1 lt lt gt gt append asin asinh atan atanh ceiling concat cons consp cos cosh eighth ensure list eq eql equal equalp evenp expt fifth find first flatten float floor format fourth intersection length list listp log max member min minusp ninth nth nth null num
32. Only roles that are not transitive and do not have any transitive subroles are allowed in number restrictions 9 Racer Systems GmbH amp Co KG http www racer systems com 58 CHAPTER 3 RACERPRO KNOWLEDGE BASES aexpr1 aexpr2 aexpr3 aexpr2 real AN AN of type real or complex real AN AN of type real aexpr3 integer AN AN of type cardinal integer AN AN of type cardinal Figure 3 3 Specific expressions for predicates DL notation RacerPro syntax At most restriction lt nR at most n R At least restriction gt n R at least n R Exactly restriction nR exactly n R Qualified at most restriction lt n R C at most n R C Qualified at least restriction gt n R C at least n R C Qualified exactly restriction n R C exactly n R C Actually the exactly restriction exactly n R is an abbreviation for the concept term and at least n R at most n R and exactly n R C is an abbreviation for the concept term and at least n R C at most n R C There are two concepts implicitly declared in every T box the concept top T denotes the top most concept in the hierarchy and the concept bottom L denotes the inconsistent concept which is a subconcept to all other concepts Note that T L can also be expressed as C UAC C N C In RacerPro T is denoted as top and L is denoted as bottom 3For KRSS compatibility reasons RacerPro also supports the synonym con
33. a lt concept expression gt gt lt role expression gt gt lt OWL datatype property gt gt see lt C gt on page 56 Fig 3 1 see lt AN gt on page 57 Fig 3 2 a RacerPro concrete domain object see lt R gt on page 56 Fig 3 1 inv lt role expressioni gt feature can be used for lt R gt as well see lt C gt on page 56 Fig 3 1 lt role expression1 gt not lt role expressioni gt see lt R gt on page 56 Fig 3 1 Note role used as datatype property p must return t true for lt R gt lt OWL annotation property gt gt see lt R gt on page 56 Fig 3 1 Note role used as annotation property p must return t true for lt R gt lt prediate expression gt gt lt CD prediate gt gt equal unequal gt lt lt gt ww min max see lt CDC gt on page 57 Fig 3 2 lt CD predicate gt string string lt gt gt lt boolean boolean lt gt divisible not divisible Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 203 6 2 The nRQL Query Processing Engine The nRQL query processing engine implements the nRQL language and is integrated with RacerPro The nRQL engine offers various querying modi In this section we describe the core functionality of this engine 6 2 1 The Query Processing Modi of nRQL nRQL can be used in d
34. a b 1 2 3 will match the query atom x a b c 123 1 2 2 3 since a b 1 2 3 implies a b c 123 1 2 2 3 note that 1 2 means 1 or 2 and c denotes the character c Since data query expressions are conjunctions in CNF the query retrieve x x age 50 is equivalent to retrieve y and x age x 50 Note that the data substrate edge query expressions have a very similar structure Since disjunctions are available in the labels as well as in the queries some flexibility is achieved this is why we call it a semi structured data model For example it is easy to retrieve the nodes whose label includes age and which are either 50 or 80 retrieve x x age 50 80 gt x betty x alice retrieve x x age 50 60 gt x betty retrieve x x age 50 Alice gt x betty x alice Note that 50 80 represents a disjunction Moreover of course also union is available Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 173 6 1 8 1 5 Using Data Query Predicates Predicates can be used in substrate query atoms So far we have only used literals Although disjunctions are available even without union the matching conditions which can be specified for nodes are somewhat limited Predicates offer much more flexibility and querying power nRQL offers so called d
35. a top constraints told value age a age a gt constraints 30 age of a firerule b top constraints told value age b age b gt constraints 60 age of b firerule and a an age b an age constraints told value age b 30 told value age a gt constraints 60 30 30 N abox consistent gt t 6 1 6 6 Using Lambda Expressions in Rule Consequences Another unique feature of the nRQL rules is their support for lambda head projection operators in rule consequences In 6 1 3 3 we showed how MiniLisp lambda expressions can be used in a query head we used the lambda ex1 racer to compute the size of a the object i The lambda ex1 racer file contains full reset define concrete domain attribute width type integer define concrete domain attribute length type integer instance i and equal width 10 equal length 20 Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 157 Implicitly we already know that i has a size of 10 20 200 however we would like to add this computed value with a rule to the ABox as well to make the value explicit racer read file nrql user guide examples lambda ex1 racer gt OKAY all attributes gt width length firerule x and a width a length instance Tx size lambda w 1 first w first 1 told value if exists
36. alice betty x eve alice betty x doris alice betty x charles Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 141 alice betty x betty alice betty x alice alice alice x eve alice alice x doris alice alice x alice These are 2 23 25 5 5 pairs Please note that the sets returned by the atom and its negated variant are complementary however if a projection is performed on a set then the sets resulting from the projection are not necessarily complementary to one another A projection is either specified explicitly by means of the project to operator or implicitly as specified by the query head retrieve x alice x has child gt x charles x betty retrieve x meg alice x has child gt x eve x doris x charles x betty x alice This seems odd since also the children of alice are returned However the result is cor rect since this is just the answer returned by retrieve alice x meg alice x has child projected to x So to retrieve the individuals which are not children of alice the following query must be used retrieve x neg project to x alice x has child gt x eve x doris x alice Note that here the neg operator is applied after the projection to x has been carried out so neg is applied to and returned a one dimensi
37. also corresponding ABox individuals are created and instance assertions are added This spatio thematic representation can now be queried with nRQL as the following exam ples demonstrates retrieve x PRY z and x and string has name Hamburg some in country germany y and country gt inhabitants 8000000 x y pp y z pp x hamburg y germany z europe Vv retrieve P x Pry and x y dr y and city some in country germany x and city some in country france gt x paris y hamburg N retrieve x and europe x ppi x city x paris x hamburg v Note that nRQL has deduced that the dr relation holds between paris and hamburg and that pp holds between hamburg and europe etc 6 1 8 3 2 OWL and the RCC Substrate The RCC substrate can be used on OWL KBs as well The RCC mirror substrate has both the functionality of the mirror data substrate as well as of the RCC substrate just discussed Thus it automatically creates and populates the substrate from the OWL ABox Moreover certain OWL object properties will be recognized as RCC relations given they have been declared as RCC synonyms in advance Consider the following OWL KB nrql user guide examples owl rcc owl Racer Systems GmbH amp Co KG http www racer systems com 190 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL lt xml version 1 0
38. and object contains object c but you also specify that a and c are not overlapping disjoint Obviously such a spatial situation can never be realized in the physical world The RCC substrate will thus become inconsistent Moreover in an RCC network substrate logically implied entailed relationships will hold These entailed relationships can be retrieved with nRQL For example suppose you construct a network in which a contains b PPI a b and in which a and c are disconnected DC a c A nRQL query asking for disconnected objects will now not only return a c but also b c since DC b c is logically entailed Usually you will not create an isolated RCC network but a network in which the nodes also have associated corresponding ABox individuals We have described this functionality when we discussed the ordinary data substrate Thus the RCC substrate can serve an ABox as an additional representation medium which is aware of the special characteristics of RCC relationships and thus can be used the answer qualitative spatial queries Let us consider the example file rcc substrate racer the s are comments full reset in tbox geo example define concrete domain attribute inhabitants type cardinal define concrete domain attribute has name type string define primitive attribute has language Racer Systems GmbH amp Co KG http www racer systems com 188 CHAPTER 6 THE NEW RACERPRO QUE
39. each other The query entailment check is complete for queries which do not contain neg The query entailment check is reduced to a series of query consistency checks The query and ai1 ain entails the query and a21 a2m iff all queries and all aln neg a21 are inconsistent moreover if the queries contain union then they are brought into DNF and each combination of disjuncts is checked accordingly Note that we have just stated that the query inconsistency checker is complete for queries containing at most one neg Thus we claim that the entailment check is complete for queries which contain no negs Moreover if should be noted that the entailment check implemented here only considers the bodies of the queries the head is irrelevant What matters for query subsumption is thus not a subsumption set inclusion relationship between the query results but between the tuple sets denoted by the query bodies Thus the query retrieve x x woman is entailed by retrieve x grandmother Obviously there is also variable renaming issue Should the queries x woman and y woman be considered as equivalent And what about and x man y woman vs and y man x woman Are they equivalent In nRQL the answer to the first question is yes and the answer to the second question is no The rationale is that we want to avoid having to consider an exponential factorial number of possible ways to rename variable
40. gt lt rdf RDF xmlns rdf http www w3 0rg 1999 02 22 rdf syntax ns xmlns http www owl ontologies com Ontology1162148702 0w1 xmlns xsd http www w3 org 2001 XMLSchema xmlns rdfs http www w3 0rg 2000 01 rdf schema xmlns owl http www w3 org 2002 07 owl xml base http www owl ontologies com Ontology1162148702 owl gt lt owl Ontology rdf about gt lt owl Class rdf ID city gt lt rdfs subClass0f gt lt owl Class rdf ID geo thing gt lt rdfs subClass0f gt lt owl Class gt lt owl Class rdf ID continent gt lt rdfs subClassOf rdf resource geo thing gt lt owl Class gt lt owl Class rdf ID country gt lt rdfs subClassOf rdf resource geo thing gt lt owl Class gt lt owl ObjectProperty rdf ID contains gt lt rdfs domain rdf resource geo thing gt lt owl ObjectProperty gt lt owl ObjectProperty rdf ID adjacent gt lt rdfs domain rdf resource geo thing gt lt owl ObjectProperty gt lt owl ObjectProperty rdf ID disjoint gt lt rdfs domain rdf resource geo thing gt lt owl ObjectProperty gt lt country rdf ID germany gt lt contains gt lt city rdf ID hamburg gt lt city gt lt contains gt lt country gt lt continent rdf ID europe gt lt contains gt lt country rdf ID france gt lt contains gt lt city rdf ID paris gt lt contains gt lt adjacent rdf resource germany gt lt country gt lt contains gt
41. gt lt rdfs domain rdf resource class gt lt owl 0bjectProperty gt Racer Systems GmbH amp Co KG http www racer systems com 182 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL lt owl DatatypeProperty rdf ID datatype property gt lt rdfs range rdf resource http www w3 org 2001 XMLSchema int gt lt rdfs domain rdf resource class gt lt owl DatatypeProperty gt lt owl ObjectProperty rdf ID annotation object property gt lt rdf type rdf resource http www w3 0rg 2002 07 owl AnnotationProperty gt lt rdfs domain rdf resource class gt lt owl ObjectProperty gt lt owl DatatypeProperty rdf ID annotation datatype property gt lt rdfs domain rdf resource class gt lt rdfs range rdf resource http www w3 org 2001 XMLSchema int gt lt rdf type rdf resource http www w3 org 2002 07 owl AnnotationProperty gt lt owl DatatypeProperty gt lt class rdf ID test gt lt rdfs comment rdf datatype http www w3 org 2001 XMLSchema string gt Comment for individual test lt rdfs comment gt lt object property rdf resource test gt lt datatype property rdf datatype http www w3 org 2001 XMLSchema int gt 123 lt datatype property gt lt annotation object property rdf resource test gt lt annotation datatype property rdf datatype http www w3 0rg 2001 XMLSchema int gt 456 lt annotation datatype property gt lt class gt lt rdf RDF gt We have
42. http www racer systems com 5 3 THE PUBLISH SUBSCRIBE MECHANISM 83 CLIENT 4 DOCUMENT 2 CLIENT 4 DOCUMENT 4 CLIENT 4 DOCUMENT 5 The query of client 4 is answered with three documents Next we discuss an example demonstrating that sometimes subscriptions do not lead to an immediate answer w r t the current A box subscribe client 2 computer science best seller The result is Although document 7 is an instance of computer science best seller this individual has already been indicated as a result of a previously subscribed query In order to continue our example we introduce two additional documents one of which is a computer science best seller state instance document 8 computer science best seller constrained document 8 isbn 8 isbn constraints equal isbn 8 8234567 instance document 9 book constrained document 9 isbn 9 isbn constraints equal isbn 9 9234567 The publish subscribe mechanism requires that these documents are published state publish document 8 publish document 9 The RacerPro system handles all publish statements within a state as a single publish statement and answers the following as a single list of subscription individual pairs CLIENT 1 DOCUMENT 9 CLIENT 2 DOCUMENT 8 CLIENT 3 DOCUMENT 8 CLIENT 4 DOCUMENT 9 Now client 2 also get information about instances of computer science best seller Note that document 8 is an instance of computer science best selle
43. o aooe oee 0002 eee ee ee 72 4 4 Closed World Assumption o osoo e 73 4 5 Unique Name Assumption LL 73 4 6 Differences in Expressivity of Query and Concept Language 73 A OWOC METAS cacra ah a ee a ee Ok A a e ai E a a Ge 74 5 Knowledge Base Management TT 5 1 Configuring Optimization Strategies 77 5 2 The RacerPro Persistency Services 0002 eee ee ee 78 5 3 The Publish Subscribe Mechanism 2 02020004 79 5 3 1 An Application Example 79 5 3 2 Using JRacer for Publish and Subscribe 84 5 3 3 Realizing Local Closed World Assumptions 85 Racer Systems GmbH amp Co KG http www racer systems com CONTENTS iii 6 The New RacerPro Query Language nRQL 87 6 1 The aRQD Languag s o ica 2 644 na bee RR A a a 93 6 1 1 Query Atoms Objects Individuals and Variables 93 6 1 2 Query Head Projection Operators 0 00002 eee 110 6 1 3 Lambda Head Operators to Evaluate Expressions 118 614A Complex Queries gt sa a a a mag poa a be Re ER was 127 61 5 Detned Queries eas ee ek e ER ee ee eS 147 6 1 6 ABox Augmentation with Simple Rules 153 6 1 7 Complex TBox Queries 2 02002 ee cin 160 6 1 8 Hybrid Representations with the Substrate Representation Layer 167 O18 Formal Syntax oF ROL ea cess p eek RR Ae Re Oe ee es 193 6 2 The nRQL Query Processing Engine 0 0 000 eee 203 6 2
44. the so called annotation properties are used to annotate resources with meta data e g comments on authorship of an ontology etc Annotation properties are not used for Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 115 reasoning they are simply ignored Nevertheless meta data is important and one should be able to retrieve this information from an OWL document OWL distinguishes two kinds of annotation properties 1 annotation object properties and 2 annotation datatype properties Fillers told values of annotation datatype properties are XML Schema datatype XSD values literals whereas fillers of annotation object properties are ordinary ABox OWL RDF individuals To retrieve a filler of an annotation datatype property spe cial head projection operators must be used similar to the datatype property case annotation datatype fillers To retrieve fillers of annotation object properties ordi nary role query atoms can be used Please note that no reasoning is performed on anno tation properties Thus no complex retrieval conditions can be specified with the help of concept query atoms However with the help of the so called mirror data substrate this becomes possible see below Let us consider an example OWL file owl ex5 ow1 document which contains all four kinds of properties lt xml version 1 0 gt lt rdf RDF xmlns http www owl ontologies com O
45. www racer systems com for the formal specification of the semantics of nRQL The semantics is based on the notion of logical entailment From a theoreticians point of view nRQL offers at least so called grounded conjunctive queries but goes beyond these by offering additional constructs Moreover the oper ational aspects of the query answering engine see below make it somehow unique nRQL only offers so called must bind or distinguished variables which range over the individuals of an ABox or individuals of an RDF S OWL document Thus a concept query atom see below such as x mother has the same semantics as concept instances mother A variable is bound to an ABox individual if and only if this individual satisfies the query Satisfies means that the query resulting from substituting all variables with their current bindings resulting in the so called grounded query is logically entailed by the knowledge base KB So x is bound to the individual betty if and only if the ground conjunctive query betty mother is entailed nRQL does not offer so called do not bind or non distinguished variables A variable is always distinguished even if the bindings of that variables bindings are not included in the query answer the query answer contains only the bindings of those variables which appear in the head of the query see below However purely existentially quantified retrieval conditions such as there exists a f
46. x doris y betty x doris y alice x betty y charles x betty y betty x betty y alice x alice y eve x alice y doris x alice y alice x betty y eve x betty y doris x alice y betty These are 24 tuples only x alice y charles is missing Indeed this tuple is returned by the non negated variant of this query if we also add y to the head retrieve x y and x woman x y has child y man gt x alice y charles This shows that the two queries are indeed complementary to one another but due to the final projection to x this no longer holds for the returned sets The solution is to first apply the projection to x and then build build a one dimensional complement set with neg as follows retrieve x neg project to x and x woman x y has child y man gt x eve x doris x charles x betty Please note that a project to body can also be understood as a kind of subquery In fact project to x and x woman x y has child y man can be understood as a subquery retrieve x and x woman x y has child y man Only after the complete result set of that subquery has been computed the com plement can be computed with neg Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 145 Please note that neg and projec
47. 0w1 xmlns daml http www daml 0rg 2001 03 daml 0i1 xmlns http ww owl ontologies com unnamed owL xmlns dc http purl org dc elements 1 1 xml base http ww owl ontologies com unnamed owl gt lt owL Ontology rdf about gt lt owL imports rdf resource http ww daml org rules proposal swrlb owl1 gt lt owL imports rdf resource http waw daml org rules proposal swrl owL gt lt owl Ontology gt lt owl Class rdf ID Person gt lt owl 0bjectProperty rdf ID hasUncle gt lt omwl 0bjectProperty rdf ID hasMother gt lt owl 0bjectProperty rdf ID hasBrother gt lt Person rdf ID J0OHN gt lt Person rdf ID SUE gt lt hasBrother rdf resource JOHN gt lt Person gt lt swrl Imp rdf ID Rule 1 gt lt swrl body gt lt swrl AtomList gt lt rdf first gt lt swrl IndividualPropertyAton gt lt swrl argument2 gt lt swrl Variable rdf ID y gt lt swrl argument2 gt lt swrl propertyPredicate rdf resource hasMother gt lt swrl argumenti gt lt swrl Variable rdf ID x gt lt swrl argumenti gt lt swrl IndividualPropertyAtom gt lt rdf first gt lt rdf rest gt lt swrl AtomList gt lt rdf rest gt lt swrl AtomList gt lt rdf first gt lt swrl DifferentIndividualsAtom gt lt swrl argumentl rdf resource x gt lt swrl argument2 rdf resource y gt lt swrl DifferentIndividualsAtomn gt lt rdf first gt lt rdf rest rdf resource http ww w3 0rg 1999 02 22
48. 333 nil master top query ig nil query 1 335 ___nil master bottom query gt see output on stdout Racer Systems GmbH amp Co KG http www racer systems com Chapter 7 Outlook Future releases of RacerPro will provide e Support for complete reasoning on SHOT Q D knowledge bases SHZQ nomi nals in concept terms arithmetic concrete domains with n ary predicats and without features chains 13 New optimization techniques for nominals have to be developed e Instead of role hierarchies more expressive role axioms so called acyclic role axioms can be supported and are useful in practice The description logic is called SRTQ 13 Future version of RacerPro will support acyclic role axioms New optimization techniques have to be developed for this language feature e Feature chains for w admissible concrete domains such as RCC or pointizable ALLEN 16 e Feature chain equality for ALCF D e SWRL rules with first order semantics e Persistency for ontologies persistency for A boxes database access e Proper algorithmic support for incremental A box changes The order in this list says nothing about priority and this list is probably not complete Racer Systems GmbH amp Co KG http www racer systems com 238 CHAPTER 7 OUTLOOK Racer Systems GmbH amp Co KG http www racer systems com Appendix A Another Family Knowledge Base In this section we present another family
49. Co KG http www racer systems com 70 CHAPTER 4 DESCRIPTION LOGIC MODELING WITH RACERPRO Note however that the latter query would not return results for the following A box whereas the former query would return i the query language uses the active domain semantics for variable bindings in knowledge base test instance i and a some r b For a detailed explanation see Chapter 6 In any case description logics and ontology lan guages such as OWL are important if concept names have definitions in the T box or if the A box contains indefinite descriptions such as instance john or french italian Although it is possible to represent an object based view of a database as an A box cur rently description logic systems do not provide for transactions and persistency of data and in order to ensure decidability in the general case the query language is in some sense less expressive than for instance relational database query languages such as SQL So mass data representing definite information is better stored in databases right now De scription logic and semantic web technology comes into play when indefinite information disjunctive information is to be treated as well See also the comment about the open world and closed world assumptions below Databases employ the closed world assumption What is not explicitly stated in the database is assumed to be false This is not the case for description logics 4 2
50. LANGUAGE NRQL 6 1 3 3 1 Aggregation Operators First we show how queries with aggregation op erators can be implemented Consider the following KB which models the compositional structure of a car A car has certain parts and each part has a certain weight see file lambda ex2 racer full reset define primitive role has part transitive t define concrete domain attribute weight type real instance carl car related cari enginel has part related engine cylinder 1 4 has part related cari wheel 1 4 has part related cari chassisi has part instance enginel weight 200 0 instance chassis1 weight 400 0 instance wheel 1 4 weight 30 0 Using MiniLisp we can compute the overall weight as well as identify its number of com ponents see file nrql user guide examples lambda ex2 queryi racer retrievet car car lambda car Let w reduce flatten retrievel and car car car part has part part a weight lambda weight weight told value if exists weight part parts length retrieve part car part has part car car no of parts parts total weight w car gt car cari no of parts 4 total weight 630 0 Please note that retrievel is like retrieve but with head and body argument positions flipped The body of the query consists of the concept query atom x car The lambda ex pression is then applied to th
51. NRQL inv http www w3 org 2002 07 owl comment inv http www w3 0rg 2000 01 rdf schema seeAlso inv http www w3 org 2002 07 owl seeAlso inv http www w3 org 2000 01 rdf schema isDefinedBy inv http www w3 org 2002 07 owl isDefinedBy inv http www w3 0rg 2000 01 rdf schema label inv http www w3 org 2002 07 owl label inv http www w3 org 2002 07 owl versionInfo http www w3 org 2000 01 rdf schema comment http www w3 org 2002 07 owl comment http www w3 org 2000 01 rdf schema seeAlso http www w3 org 2002 07 owl seeAlso http www w3 org 2000 01 rdf schema tisDefinedBy http www w3 org 2002 07 owl isDef inedBy http www w3 org 2000 01 rdf schema label http www w3 org 2002 07 owl label http www w3 org 2002 07 owl versionInfo all individuals gt http www owl ontologies com unnamed owl boo0k123 http www owl ontologies com unnamed owl michael Alternatively you can also use the RacerPorter Concepts Roles and Individuals Tabs in RacerPorter and turn off the Simplify Names checkbox nRQL is more powerful than other RDF S query languages for example true classical disjunction is available retrieve x x or book person gt x http www owl ontologies com unnamed owl boo0k123 x http www owl ontologies com unnamed owl michael 6 1 1 2 Role Query Atoms The second type of query atoms are the role query atoms Rol
52. RULE LANGUAGE 47 K SKORS RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox temp 1 ow ABox temp 1 ow l Concept Role Individual MARY Query Rule v Simplify Names Command GSWRL FORWARD CHAINING Ji Disconnect Profiles Refresh ClearLog FullReset New Open Load RacerPro Log Erroneous cardinality specification 1 handled generously a Erroneous cardinality specification 1 handled generously done r Concept http www w3 org 2003 11 swrl AtomList causes a cycle in TBox temp temp 1 owl Define http www owl ontologies com unnamed owl Rule 11 hasUncle x 2 lt hasMother x y hasBrother y 2 done 6 gt temp temp 1 owL 0 Figure 2 30 Shell window used for starting the rule engine For demonstration purposes we start the rule engine by typing a command into the Rac erPorter shell window Figure 2 30 RacerPro offers a forward chainer that applies SWRL rules until no new information is added Racer Systems GmbH amp Co KG http www racer systems com 48 CHAPTER 2 USING RACERPRO is RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox temp 1 0w1 ABox temp 1 0wl Concept Role Individual MARY Query Rule i M simplify Names Command I
53. This license is valid forever This is RacerPro for Ralf Moeller HTTP service enabled for http localhost 8686 TCP service enabled for http localhost 8688 Figure 2 8 A screenshot of the RacerPro server started under Windows Depending on the arguments provided at startup the RacerPro executable supports differ ent modes of operation It offers a file based interface a socket based TCP stream interface and a HTTP based stream interface With the OWL QL server comes a web service inter face 2 2 1 The File Interface If your knowledge bases and queries are available as files use the file interface of RacerPro i e start RacerPro with the option f In the following we assume that you have RacerPro Racer Systems GmbH amp Co KG http www racer systems com 2 2 THE RACERPRO SERVER 19 on your search path In your favorite shell on Unix based systems just type RacerPro f family racer q family queries Under Windows you want to suppress the display of the RacerPro window The command is slightly different RacerPro c f family racer q family queries lisp The option separates window management options from RacerPro options A window option we use here is c which suppress the display of the RacerPro window which is not useful in batch mode For debugging under Windows the window option p is useful RacerPro p f family racer q family queries lisp If an error occurs you can read
54. You can also dump all substrates store all substrates Restoring substrates is easy as well restore substrate restore all substrates Please consult the Reference Manual for more details on these API functions Dumping a substrate will always automatically dump the associated ABox and TBox as well Moreover there can be defined queries see Section 6 1 5 associated with a TBox the substrate can also have an associated QBox etc All these data structures are also saved However it is not possible to resurrect the queries from the QBox they merely serve as caches Thus you cannot call reexecute query on a query Id which you see in a restored QBox This list of queries processed queries active queries etc are NOT saved If you really want to save queries into a file we ask you the define these queries see Section 6 1 5 Restored definitions can be reused See the example below We have also discussed in Section 6 1 8 that there are specialized types of substrates avail able tailored for special representation tasks For example the data substrate or the RCC substrate These substrates can be saved as well and thus the created hybrid representation can be preserved The following session demonstrates the utility of the nRQL persistency facility full reset okay full reset Vv racer read file home mi wessel nrql user guide examples family no signature racer gt OKAY N enable query repository oka
55. Zal Preferences occie adc ai e ke eR wee es 242 Racerkditor 44 5 0 4 a a dea be Gage eee aa e bees 24 3 Was ta RaserPorter i ca oraaa dA bo A wee Ae a a DADAM oon A A N 18 18 20 22 22 23 24 25 26 27 28 Racer Systems GmbH amp Co KG http www racer systems com ii CONTENTS 244 Known Problems oc oces cs 688864 458244 e Re 29 2 5 Other Graphical Client Interfaces for RacerPro 0 0 29 Boel RICE 4 be eee ea ee ba ae ew ee RR oe A ee kw 29 20 2 Protege sa opos goai ma a Ee EA eo eS 31 2 5 3 Using Prot g and RacerPorter in Combination 38 2 5 4 TopBraidComposer i 43 2 6 SWRL Semantic Web Rule Language 2204 44 3 RacerPro Knowledge Bases 55 3 1 Naming Conventions s lt o sa osora na be he ee ee eS 55 3 2 Concept Language 1 a 56 33 Concept Axioms and T boxes s o sie 626 240400 es SLSR A iboi 59 3 4 Role Declarations sa s sa sste i eR eG ee 60 Bo Concrete Domains so socos kh ew ee ee oe DR Ae eR a 61 3 6 Concrete Domain Attributes 2 se cee ee bee be ee ee ee 65 3 7 Individual Assertions and A boxes LL 65 ove InterenceIMOGES ed Le Rie PCR le Ble ee Be ee Be a 66 3 9 Retraction and Incremental Additions 67 4 Description Logic Modeling with RacerPro 69 4 1 Representing Data with Description Logics 69 4 2 Nominals or Concrete Domains 0 00002 00 70 4 3 Open World Assumption
56. a ___ b woman 333 ___master bottom query gt see output on stdout query equivalents last gt NIL retrieve y and x mother y x has parent y man gt y charles show qbox for abox smith family t 333 QBOX FOR racer dummy substrate FOR ABOX smith family master top query as and x ano1 woman x anol y has child y human sigs ___ and x ano1 mother y x anoi has parent y man Sa ll ___master bottom query o x top is ___ a human 335 ___ b woman BHE ___master bottom query gt see output on stdout query children last Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 229 gt master bottom query show qbox for abox smith family nil 333 QBOX FOR racer dummy substrate FOR ABOX smith family 333 O master top query 333 4 query 9 333 ___5 query 43 tra Al ___0 master bottom query is 1 query 1 DEE ___2 query 2 SUBQUERY 3 0F query 9 333 ___3 query 5 SUBQUERY 1 OF query 9 333 ___0 master bottom query 17 gt see output on stdout query children query 9 gt query 43 query parents query 43 gt query 9 retrieve nil and a mother a b has child b man gt t show qbox for abox smith family t 333 QBOX FOR racer dummy substrate FOR ABOX smith family master top query Spa and x anoi woma
57. already listed the markers which are added for the various OWL elements Lets us briefly inspect the node and edge labels full reset okay full reset Vow N enable data substrate mirroring okay data substrate mirroring enabled Vv N owl read file nrql user guide examples owl ex5 owl nrql user guide examples owl ex5 owl Vv N describe all nodes node name 456 node label owl object owl literal 456 node successors NIL node predecessors http www owl ontologies com Ontology1159352693 owl test Vv node name http www owl ontologies com Ontology1159352693 owl class node label owl class Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 183 node successors This is a test class node predecessors NIL node name This is a test class node label owl object owl literal This is a test class node successors NIL node predecessors http www owl ontologies com Ontology1159352693 owl class node name Comment for individual test node label owl object owl literal Comment for individual test node successors NIL node predecessors http www owl ontologies com Ontology1159352693 owl test node name 123 node label owl object owl literal 123 node successors NIL node predecessors http www owl ontologies com Ontology1159352693 owl test
58. another unique feature of nRQL The semantics is analog to classical negated concepts For example RacerPro can prover that male persons can never be instances of the concept mother in family racer since mother implies female and male and female are declared as disjoint in family racer retrieve x x not woman gt x charles A negated role works analogously to a negated concept Note that negated roles are only available in the nRQL query language but not in the RacerPro concept language to grant decidability A negated role not has sister allows us to verify that male persons can never be fillers of the has sister role since family racer contains the axiom implies top all has sister some has gender female This means that fillers of the has siser role have a female gender and thus they cannot be male This matches our intuition and indeed using a negated role not has sister we see that charles cannot be a sister of anyone retrieve x y x y not has sister gt x eve y charles x doris y charles x charles y charles x betty y charles x alice y charles Please note that queries involving negated roles are rather expensive 6 1 1 2 3 Querying OWL KBs with Role Query Atoms In the OWL realm so called object properties are the equivalent of roles Consider the following OWL document owl ex2 owl in the nrql user guide examples fold
59. associated tbox b b instance b b gt OKAY defquery test x x b gt test retrieve x x test gt x b describe all definitions gt defquery test x x b defquery test2 x x top tbox a gt test2 describe all definitions Racer Systems GmbH amp Co KG http www racer systems com 152 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL gt defquery test x x b describe all definitions tbox a gt defquery test2 x x top defquery test x x a retrieve x substitute test2 x gt ERROR NRQL ERROR Can t find definition test2 in DBox for TBox b in tbox a gt a in abox a gt a describe all definitions gt defquery test2 x x top defquery test x x a retrieve x substitute test2 x gt x a Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 153 6 1 6 ABox Augmentation with Simple Rules nRQL offers a simple ABox augmentation mechanism rules nRQL rules are termination safe since there is no automatic rule application strategy the application strategy is com pletely under control of the client application user However nRQL supports rule API functionality which will be needed for the client side implementation of a rule application strategy see below 6 1 6 1 Terminology Rule Antecedent and
60. bar are strings 123 3 is a floating point number and foobar is a symbol Unlike boolean formulas data literals are always positive and thus cannot be negated Racer Systems GmbH amp Co KG http www racer systems com 168 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 1 8 1 2 Data Substrate Nodes Edges and Labels To populate a data substrate nRQL s substrate functionality must be enabled first At its creation a data substrate is associated with a certain ABox in this case the default ABox full reset gt okay full reset in data box default default v describe substrate abox default type data substrate abox default no of nodes 0 no of edges 0 v The output of describe substrate shows us that there are yet no nodes or edges in that data substrate We can now create a node in the substrate as follows data node betty Betty age 50 and woman age 50 some has child human gt betty This does three things 1 Creates a data substrate node named betty 2 labels this node with the description Betty age 50 which is a conjunction of three data literals 3 creates an ABox individual called betty and finally 4 adds the concept assertion instance betty and woman age 50 some has child human to the default ABox associated with this data substrate Note that not only symbols are valid as names for data substrate nodes but also num
61. bases in a distributed system can cause security problems The RacerPro server executes statements as described in the sections below Statements that might cause security problems are save tbox save abox and save kb Files may be generated at the server computer By default these functions are not provided by the RacerPro server If you would like your RacerPro server to support these features startup RacerPro with the option u for unsafe You also have to start RacerPro with the option u if you would like to display or edit OWL files with RacerEditor and use the facility Evaluate OWL File e If RacerPro is used in the server mode the option init lt filename gt defines an initial file to be processed before the server starts up For instance an initial knowledge base can be loaded into RacerPro before clients can connect e The option n allows for removing the prefix of the default namespace as defined for OWL files See Chapter 4 7 for details e The option t lt seconds gt allows for the specification of a timeout This is particularly useful if benchmark problems are to be solved using the file interface Racer Systems GmbH amp Co KG http www racer systems com 24 CHAPTER 2 USING RACERPRO The option debug is useful for providing bug reports If an internal error occurs when RacerPro is started with debug a stack backtrace is printed See Section 2 3 about how to send a bug report With una you can fo
62. binding for the variables in the body such that the predicates either unary or binary predicates are possible in SWRL are satisfied then the predicate comprising the head also holds The predicates in the body are also called the precondition and the head is the consequence Variables are bound to a finite set of individuals namely those explicitly mentioned in the ontology or A box The question is whether given a specific binding for variables the precondition has to be satisfied in one world or in all worlds We call the former semantics the first order semantics whereas the latter is called the rule semantics Racer Systems GmbH amp Co KG http www racer systems com 2 6 SWRL SEMANTIC WEB RULE LANGUAGE 51 e080 french italian Prot g 3 1 1 file Applications Protege_3 1 french italian pprj OWL Files owl or rdf Delaere eeN EK prot g PI Properties Forms lt Individuals lt j Metadata SP SWRL Rules SUBCLASS RELATIONSHIP CLASS EDITOR PF TT For Project french italian For Class French or Italian instance of owl Class Asiatial Rieck Gori SameAs DifferentFrom LJ Annotations Property sm French or Itaim O swrl Variable rdfs comment gt swrl Atom swrl Builtin French Italian French or Italian No wine drinker Asserted Conditions Dele x NECESSARY amp SUFFICIENT French v Italian NECESSARY
63. clients see also the documentation about the RacerPro Proxy For instance you can classify a T box or realize an A box and dump the resulting data Racer Systems GmbH amp Co KG http www racer systems com 5 3 THE PUBLISH SUBSCRIBE MECHANISM 79 structures into a file The file s can be reloaded and multiple servers can restart with much less computational resources time and space Starting from a dump file is usually about ten times faster than load the corresponding text files and classifying the T box or realizing the A box again Since future versions of RacerPro might be supported by different internal data structures it might be the case that old dump files cannot be loaded with future RacerPro versions In this case an appropriate error message will be shown However you will have to create a new dump file again If you have a license for RacerMaster dumping an image is possible with the underlying Common Lisp technology and is much much faster 5 3 The Publish Subscribe Mechanism Instance retrieval see the function concept instances is one of the main inference services for A boxes However using the standard mechanism there is no efficient way to declare so called hidden or auxiliary individuals which are not returned as elements of the result set of instance retrieval queries Furthermore if some assertions are added to an A box a previous instance retrieval query might have an extended result set
64. connected to RacerPro and some knowledge base verifications on pizza owl a Prot g example knowledge base are to be performed In Figure 2 20 you can seen the pizza ontology loaded into Prot g Ontology verification is started by pressing the icon labeled with C classify The results are shown in Figure 2 21 Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO 39 file Applications Protege_3 1 examples pizza pizza owl pprj OWL Files owl or rdf ERAY eee pizza owl Prot g 3 1 1 E C E PE E Properties Forms lt I gt Individuals lt gt Metadata LASS EDIT For Project pizza owl For Class Capricciosa instance of owl Class Asserted Hierarchy Name SameAs DifferentFrom D Annotations Value owl Thing Property Y DomainConcept Capricciosa IO pyrdtstabel Capricciosa county cs F IceCream v Pizza CheeseyPizza InterestingPizza MeatyPizza Y NamedPizza American sai AmericanHot __ Asserted Inferred Properties UG x Cajun zr gt O hasBase single PizzaBase Capricciosa Asserted Conditions GCH v hasTopping multiple PizzaTopping Caprina a5 NECESSARY amp SUFFICIEN Toma
65. consult the API description of the data substrate facility in the Reference Manual 6 1 8 1 3 Querying the Hybrid Representation nRQL can now be used to query the hybrid ABox Substrate representation A hybrid query consists of atoms of two kinds atoms for the data substrate and atoms for the ABox The reader is already familiar with the ABox query atoms The data substrate atoms are introduced now Both kinds of atoms are combined using the already introduced query body constructors and union neg project to So a hybrid query consists of two kinds of atoms In order to distinguish substrate from ABox atoms a simple idea is used The objects variables individuals in the substrate atoms are simply prefixed with an asterix in order to distinguish them from the atoms which reference the ABox retrieve Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 171 x x y y and x woman x y has child x Alice y age 50 gt x alice x alice y betty y betty So the variables prefixed with are bound to substrate nodes and the remaining variables to ABox individuals This means that x woman and x y has child are ABox query atoms and x Alice and y age 50 are data substrate query atoms We will discuss these new atoms in a minute The x prefixed variables are called substrate variables see rule lt data substrate query object
66. dants y and z that both have u as a descendant Moreover we must ensure that y and z are not tbox re x and x x y z neg neg neg 7 gt x Cex x x x x x x x Cox x x So nRQL h on the same path i e z is neither a descendant nor an ancestor of y trieve y z u y has descendant z has descendant u has descendant fu has descendant same as y z y z has descendant y z has ancestor woman y mother z woman with children u bottom woman y mother z woman with children u bottom woman y woman with children z mother u bottom woman y woman with children z mother u bottom top y mother z woman with children u bottom top y mother z woman with children u bottom top y woman with children z mother u bottom top y woman with children z mother u bottom top y mother z woman with children u bottom top y mother z woman with children u bottom top y woman with children z mother u bottom top y woman with children z mother u bottom as identified the synonym concepts mother and woman with children as pos sible bindings for y and z However we want to exclude synonym nodes here tbox re trieve Racer Systems GmbH amp Co KG http www racer systems com 166 CHAPTER 6 THE NEW R
67. edu X7Ezhp2 2004 0401 univ bench owl memberOf y http www University0 edu Inttp www lehigh edu 7Ezhp2 2004 0401 univ bench owl subOrganization0Of Racer Systems GmbH amp Co KG http www racer systems com 257 x y benchmark universities departments 3 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Student y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Faculty z http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Course x y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl advisor x z http www lehigh edu X7Ezhp2 2004 0401 univ bench owl takesCourse y z http www lehigh edu X7Ezhp2 2004 0401 univ bench owl teacher0Of x y z defun run lubm tests benchmark function univs amp key max deps univ 1 nil mode 1 check abox consistency gt mode 2 full reset set nrql mode mode enable optimized query processing mode 1 set unique name assumption t when and gt univs 1 max deps univ 1 error Maximum number of departments may only be specified if only one university is processed run lubm benchmark benchmark function check abox consistency 1 univs and max deps univ 1 1 max deps univ 1 defun testi n amp optional mode 1 check abox consistency gt mode 2 run lubm tests simple lubm n check abox consistency check abox consistency mode mode defun test2 n amp o
68. existence of some has child filler in the ABox whereas for query 6 this child most not be explicitly present in the ABox However an alternative definition of mother will be recognized as equivalent to the query 6 not that Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 225 prepare abox query nil and x and woman some has child top gt query 9 ready to run query entails p query 9 query 6 gt t query entails p query 6 query 9 prepare abox query x x mother gt query 12 ready to run prepare abox query nil x grandmother gt query 13 ready to run query entails p query 13 query 12 gt t prepare abox query nil and x woman y man gt query 16 ready to run prepare abox query nil and a woman b man gt query 17 ready to run prepare abox query nil and b woman a man gt query 18 ready to run query equivalent p query 16 query 17 gt t query equivalent p query 16 query 18 gt NIL prepare abox query nil and a mother b uncle c grandmother gt query 22 ready to run N query entails p query 22 query 16 gt t Racer Systems GmbH amp Co KG http www racer systems com 226 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL query entails p query 16 query 22 gt NIL
69. explicitly stated the color of traffic light 2 is not red Thus only green remains and due to the other constraints the colors of the other traffic lights are determined It is obvious that instead of string values one could have used nominals However optimized reasoning algorithms for nominals will only be part of a future version of RacerPro Racer Systems GmbH amp Co KG http www racer systems com 72 CHAPTER 4 DESCRIPTION LOGIC MODELING WITH RACERPRO Using nominals might be even tricky Consider the following knowledge base equivalent shabby car all has color shabby color equivalent reddish object some has color one of orange red instance car 1 and reddish object not some has color one of orange instance car 2 and reddish object not some has color one of orange Now assume some information of the color of car 1 is available car 1 is an old car instance car 1 all has color shabby color It is obvious that due to the use of nominals car 2 is a shabby car as well This is probably unintended and is most likely a modeling error 4 3 Open World Assumption As other description logic systems RacerPro employs the Open World Assumption OWA for reasoning This means that what cannot be proven to be true is not believed to be false Given the T box and A box of the family example see the previous chapters a standard pitfall would be to think that RacerPro is wrong considering
70. family knowledge base is presented in Figure 2 10 The query searches for instances of the concept or man woman Other queries e g as those shown in the previous section can be submitted to RacerPro in a similar way An example for submitting a statement is displayed in Figure 2 11 The family knowledge base is exported as an OWL file Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO 31 amp RACER Interactive CLIENT Environment RICE File Edit Tools Help Concepts Type a Concept Name to show exact matching only ABoxes p DEFAULT E PERSON Ey MAN N BROTHER 1 UNCLE on FATHER E PARENT Ei N WOMAN W MOTHER EN SISTER Son AUNT soon MALE Statements and or cancept definitions save kb CJRalffKBs family kb dam thox family abox smith family syntax daml ins file C RalffKBs t PreClassify RACER Replies OK Figure 2 11 Screenshot of RICE 2 5 2 Prot g Prot g may be used as another graphical interface for RacerPro Prot g is available from http protege stanford edu While RICE can be used to pose queries in particular for A boxes Prot g can be used to graphically construct T boxes or ontologies and A boxes Racer Systems GmbH amp Co KG http www racer systems com 32 CHAPTER 2 USING RACERPRO 600 family 1 Prot g 3 1 beta file Users rm Desktop family 1 ppr
71. features e The complete API of the nRQL engine is specified in the RacerPro Reference Manual Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 93 6 1 The nRQL Language In the following we introduce the nRQL language bottom up in a tutorial like style Bottom up means that we introduce the language by first considering the various available query atom and then introduce complex compound queries build from these atoms Please note that the example files we are using in this section are con tained in the subfolder called nrql user guide examples in the examples archive which can be downloaded from your personal download page under http www racer systems com Moreover it is assumed that you use RacerPorter for the tutorial interaction with RacerPro In case you get different results as listed here please push the button Full Server Reset use the Server or Shell Tab and retry the interaction 6 1 1 Query Atoms Objects Individuals and Variables The basic expressions of the nRQL language are so called query atoms or simply atoms In this Subsection we are only considering ABox query atoms nRQL offers TBox and Substrate atoms as well see below Atoms are either unary or binary A unary atom references one object and a binary atom references two objects An object is either an individual or a variable In case of an ABox
72. for answering queries For instance in order to answer a query for the parents of a concept the T box must be classified However for answering an instance retrieval query this is not necessary and therefore RacerPro does not classify the T box in the standard inference mode Nevertheless if multiple instance retrieval queries are to be answered by RacerPro it might be useful to have the T box classified in order to be able to compute an index for query answering Considering a single query RacerPro can not determine whether computing an index is worth the required computational resources Therefore RacerPro can be instructed about answering strategies for subsequent queries e In order to ensure a T box is classified you can use the statement classify tbox amp optional tbox name e To compute an index for fast instance retrieval for an A box you can call compute index for instance retrieval amp optional abox name Comput ing an index is usually a costly process so this should be done offline After having computed an index for an A box one might use the persistency services of Racer Pro to dump the internal A box data structures for later reuse The index compu tation process is also known as A box realization and you can also call the function realize abox amp optional abox name to achieve the same effect e The function prepare abox can be used to compute index structures for an A box You can call this function offline to
73. format in a public available command language anybody with knowledge about the existence of a RacerPro server can manipulate the server status the data stored and the processing since it is also possible to connect to one RacerPro server instance from multiple clients which may not be all under your control 2 2 2 TCP APIs There are two APIs based on TCP sockets namely for the programming languages Common Lisp and Java The socket interface of the RacerPro server can be used from application programs or graphical interfaces Bindings for C and Prolog dialects have been developed as well If the option f is not provided the socket interface is automatically enabled Just execute the following RacerPro The default TCP communication port used by RacerPro is 8088 In order to change the port number the RacerPro server should be started with the option p For instance RacerPro p 8000 or RacerPro p 8000 under Windows In this document the TCP socket is also called the raw TCP interface The functionality offered by the TCP socket interface is documented in the next sections RacerPro s server engine listens on the specified TCP port for incoming connections from client processes running on the local machine the network or even the internet Please note that RacerPro itself offers no fraud and privacy protection You have to utilize network based security services such as firewalls IP tunnels VLAN switches and simil
74. function sleeping queries returns this list The function query sleeping p is used to check whether a query is on this list The thread of a query which appears on this list is currently sleeping the query had been started in lazy incremental mode Note that queries started in the set at a time mode will never appear on this list 4 The list of terminated inactive queries The corresponding API func tion terminated queries also processed queries this list The function query terminated p also query processed p is used to check whether a query is on this list A query is put on this list when its query answering thread terminates Note that different events can cause a query answering thread to terminate e g all query answer are computed the query is aborted the timeout is reached etc The function reprepare query puts an already processed query back on the list of ready queries and so a fresh life cycle can begin reincarnation It is also possible to directly re execute and thus put a query back on the list of active queries by using the API function reexecute query Note that this function only applies to terminated queries not to ready queries However there is the function execute or reexcute all queries 6 2 3 The Life Cycle of a Rule The same six different lists are also maintained for nRQL rules RacerPorter provides the Rules tab which can be used to inspect and manage the rules as well as their curren
75. gt Page 193 Moreover the variables in these atoms are bound pairwise in parallel If x is bound to alice then x is automatically bound to the corresponding substrate node alice and vice versa In case there is no corresponding node or individual the pairwise binding fails and the query returns NIL resp FALSE Query atoms can reference individuals resp substrate nodes as well The same naming convention is used to refer to substrate nodes from within query atoms Substrate nodes are prefixed with as well Considering the result delivered by the query above we see that x alice and x alice even though the name of the substrate node is in fact alice retrieve nil alice Alice gt t retrieve nil alice Alice gt ERROR NRQL ERROR Parser Error Unrecognized concept expression Alice Not only the substrate nodes can be queried but also the edges Let us add one more atom a so called data substrate edge query atom retrieve x x y y and x woman x y has child x Alice y age 50 x y has child gt x alice x alice y betty y betty There are also injective data substrate variables All variables beginning with are injec tive data substrate variables retrieve x y and top x top y gt x alice y alice x alice y betty x betty y alice Racer Systems GmbH amp Co KG http www
76. interface is automatically enabled If you do not use the HTTP interface at all but the TCP interface only start RacerPro with the option http 0 Clients can connect to the HTTP based RacerPro server using the POST method For details see the DIG standard 4 The default HTTP communication port used by RacerPro is 8080 In order to change the port number the RacerPro server should be started with the option http For instance RacerPro http 8000 under Unix and RacerPro http 8000 under Windows Console logging of incoming POST requests in provided by default but can be switched off using the option nohttpconsolelog With the option httplogdir lt directory gt logging into a file in the specified directory can be switched on The DIG standard as it is defined now is just a first step towards a communication standard for connecting applications to DL systems RacerPro provides many more features that are not yet standardized These features are offered only over the TCP socket interface However applications using RacerPro can be developed with DIG as a starting point If other facilities of RacerPro are to be used the raw TCP interface of RacerPro can be used in a seemingless way to access a knowledge base declared with the DIG interface If later Racer Systems GmbH amp Co KG http www racer systems com 2 2 THE RACERPRO SERVER 23 on the standardization process make progress users should be able to easily ada
77. its answer to the following query individual instance alice at most 2 has child RacerPro answers NIL However NIL does not mean NO but just cannot be proven w r t the information given to RacerPro Absence of information w r t a third child is not interpreted as there is none this would be the Closed World Assumption CWA It might be the case that there will be an assertion related alice william has child added to the A box later on Thus the answer NIL is correct but has to be interpreted in the sense of cannot be proven Note that it is possible to add the assertion instance alice at most 2 has child to the A box Given this the A box will become inconsistent if another individual e g william is declared to be a child of alice Many users asked for a switch such that RacerPro automatically closes roles However this problem is ill defined A small example should suffice to illustrate why closing a role or even a KB is a tricky problem Assume the following axioms disjoint a b c instance i and some r a some r b some r c some r d related i j r Racer Systems GmbH amp Co KG http www racer systems com 4 4 CLOSED WORLD ASSUMPTION 73 Now assume the task is to closed the role r for the individual i Just determining the number of fillers of r w r t i and adding a corresponding assertion at most 1 r to the A box is a bad idea because the A box gets inconsistent Du
78. knowledge base is consisting of a T box and an A box The conceptual knowledge is represented in the T box and the knowledge about the instances of a domain is represented in the A box For more information about the description logic SHIQ supported by RacerPro see 9 The extension of expressive description logics with concrete domains is discussed in 6 3 1 Naming Conventions Throughout this document we use the following abbreviations possibly subscripted C Concept term name Name of any sort CN Concept name S List of Assertions IN Individual name GNL List of group names ON Object name LCN List of concept names R Role term abox A box object RN Role name tbor T box object AN Attribute name n A natural number ABN A box name real A real number TBN T box name integer An integer number KBN knowledge base name string A string The API is designed to the following conventions For most of the services offered by RacerPro macro interfaces and function interfaces are provided For macro forms the T box or A box arguments are optional If no T box or A box is specified the value of current tbox or current abox is taken respectively However for the functional counterpart of a macro the T box or A box argument is not optional For functions which do not have macro counterparts the T box or A box argument may or may not be optional For RacerMaster or LRacer users All names are Lisp symbols the concepts are symbols or lists Ple
79. one another after the projec tion to x has been performed So neg x y has child returns the set of 21 pairs shown in the query retrieve x y neg x y has child above The reader can verify that we get eve doris charles betty alice if this set is projected to its first components If we want to ensure that also the query results sets constructed by the final projection the projection to x as specified by the query head of these queries are complementary we must make sure that the complement operator is applied to a set of instances and not to a set of pairs The project to operator is provided for this purpose see below retrieve x neg project to x x y has child gt x eve x doris x charles This query now solves the specified retrieval task retrieve all people without known chil dren Analog as for concept query atoms it is the case that x y not R always returns a subset of neg x y R for an arbitrary role R Moreover classical negation and NAF negation can be mixed like for concept query atoms Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 137 retrieve x y meg x y not has sister gt x eve y eve x x x x x x x x x x x x x x x x x x x eve y doris eve y betty eve y alice doris y eve dor
80. owl ontologies com Ontology1162148702 owl france get data edge description http www owl ontologies com Ontology1162148702 owl france http www owl ontologies com Ontology1162148702 owl germany gt from node http www owl ontologies com Ontology1162148702 owl france to node http www owl ontologies com Ontology1162148702 owl germany rcc relation ec edge label abox relationship abox role relationship owl relationship owl object property relationship http www owl ontologies com Ontology1162148702 owl adjacent Please note that the RCC synonym mechanism also applies to KBs ABoxes which are not in OWL format Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 193 6 1 9 Formal Syntax of nRQL Here we give an EBNF syntax definition of nRQL means zero or more occurrences X denotes a literal X Y Z means chose exactly one from the given literals a Ga uy nZ For example the list x betty age y is a valid lt abox query head gt if age is a concrete domain attribute see lt cd attribute name gt as well as x betty betty y http a com ontology a nil or t are invalid lt abox query head gt s The top level syntax is defined as follows retrieve lt query head gt lt query body gt retrieve under premise lt query premise gt lt abox query head gt lt abox query body gt tbox retrieve lt
81. racer systems com 172 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL x betty y betty retrieve x y and top x top y gt x alice y betty x betty y alice 6 1 8 1 4 Data Substrate Query Atoms A data substrate variable is bound to a substrate node having a certain node label if its label satisfies the given query expression We distinguish data node query expressions and data edge query expressions Both are commonly called query expressions Query expressions have a structure very similar to the structure of the labels We already said that labels are positive boolean formulas in CNF Query expressions are CNF formulas with variables Consider the node query expressions age 50 used in the data substrate query atom y age 50 in the example query above Intuitively this atoms matches all nodes which have age 50 on their node labels As for the usual ABox concept query atoms a notion of logical entailment is employed for the matching process A data node matches a given data substrate node query atom iff the label of this node logically implies the node query expression in the query atom So betty matches the query atom y age 50 because the label of betty is given as Betty age 50 resp Betty A age A 50 and thus logically implies the weaker query expression age 50 resp age A 50 in signs Betty A age A 50 age A 50 Similarly the node description
82. rdf syntax ns ni1 gt lt swrl AtomList gt lt rdf rest gt lt rdf first gt we lt swrl IndividualPropertyAtom gt X Finished evaluating localhost 8088 uncle owl XML Users rm Desktop Work SWRL Presentat Figure 2 28 RacerEditor showing the OWL SWRL code of the example Racer Systems GmbH amp Co KG http www racer systems com 46 CHAPTER 2 USING RACERPRO e SKS RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox temp 1 owl ABox temp 1 0wl Concept Role Individual MARY Query Rule Simplify Names pan hasMother gt SUE hasBrother 39N Tree Depth 1 E O Tree A Graph Transitive _ Selected Roles Roles Tab Horizontally O Vertically Focus 3 4 Reset 3 Clear Log _ Direct Types 3 All Types 4 Consistent 3 lt Realize RacerPro Log http ww w3 0rg 2003 11 swrl AtomList causes a cycle in TBox temp temp 1 owl Define http www owl ontologies com unnamed owl Ru a le 11 pa hasUncle x z lt hasMother x y hasBrother y 2 done 6 gt temp temp 1 owL 0 O v Figure 2 29 Graphical display of the instances and their relations In Figure 2 29 we use the network inspector of RacerPorter to have a look at the instances and their relations properties Racer Systems GmbH amp Co KG http www racer systems com 2 6 SWRL SEMANTIC WEB
83. roles using the tabs Taxonomy and Role Hierarchy respectively see Figures 2 4 and 2 5 e 00 RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox FAMILY ABox SMITH FAMILY Concept Role Individual Query Rule Simplify Names CFATHER CPARENT MOTHER GRANDMOTHER HUMAN PERSON MAN BROTHER UNCLE c Tope me WOMAN Z CSISTER AUNT FEMALE MALE Tree Depth 1 xs O Tree Graph C Bottom Horizontally Vertically Focus Reset Clear Log _ Concept Query Coherent Classify RacerPro Log 7 Using Profile Localhost to Connect to localhost 8088 A gt 0KAY vw Using Profile Localhost to Connect to Localhost 8 88 DKAY Figure 2 4 RacerPorter showing the taxonomy of the T box family You should select the button graph in order to see the full hierarchy You can switch between the shell the taxonomy and the role hierarchy by selecting the corresponding tabs in the RacerPorter window Next we will use RacerPorter to specify some queries and inspect the answers You can use the Shell tab or alternatively you can use the Console tab to execute queries as we have done in Figure 2 6 Information about commands and key abbreviations is printed into the Shell window With meta p or alt p and meta n or alt n you can get the previous and next command of the command history respectively You migh
84. save computational resources at query answering time In addition you can call prepare racer engine to compute index structures for query answering Again this function is to be used offline to save computational resource for answering the first query e If multiple queries are to be answered and each query will prob ably be more specific than previous ones use the directive enable subsumption based query processing The T box is then classified once the first query is answered Racer Systems GmbH amp Co KG http www racer systems com 78 CHAPTER 5 KNOWLEDGE BASE MANAGEMENT e If you use nRQL it is possible to instruct RacerPro to use less costly algorithms for query answering see also Chapter 6 This can be done by calling set nrql mode 1 In this case query answering is complete only for hierarchies T boxes with very simple axioms If for every exists restrictions declared in the T box there exists an explicit filler declared in the A box you can instruct RacerPro to speedup query answering even more by calling enable optimized query answering Note that it is necessary that you call enable optimized query answering before you read the OWL files comprising the knowledge base e The function abox consistent can be explicitly called Before the first instance retrieval query is answered the A box in question is checked for consistency You might want to call this function offline to save computational resourc
85. tbox query head gt lt tbox query body gt defquery lt query name gt lt def query head gt lt query body gt firerule lt rule antecedence gt lt rule consequence gt Syntax of the query heads and premises lt query premise gt gt List of RacerPro ABox assertions lt query name gt gt lt symbol gt naming a defined query lt symbol gt gt any Lisp symbol e g huhu foobar http a com ontoly a WITH THE FOLLOWING EXCEPTIONS all types types instantiators all instantiators all types flat all instantiators flat types flat instantiators flat direct types most specific types most specific instantiators direct instantiators direct types flat most specific types flat most specific instantiators flat direct instantiators flat describe individual synonyms concept synonyms describe nrql equal role bind individual inv not neg and or one of racer satisfies top bottom true query false query constraint has known successor substitute insert same as equal intersection union cap cup lambda told value told values told value if exists told values if exists datatype filler datatype fillers annotation datatype filler annotation datatype fillers unknown operator bindings from no told value no known cd objects project project to pi real top real bottom has child has parent has descendant has ancestor has synonym lt def query head gt gt lt def head entry gt lt abox query head gt gt lt abox head entry gt
86. that we have not yet introduced the neg and and operators that appear in the following example but an intuitive understanding is sufficient for the moment full reset okay full reset v instance santa claus good man gt OKAY instance weihnachtsmann good man gt 0KAY retrieve x y and x good man y good man gt x santa claus y santa claus x santa claus y weihnachtsmann x weihnachtsmann y santa claus x weihnachtsmann y weihnachtsmann retrieve x y and x good man y good man same as x y gt x santa claus y santa claus x weihnachtsmann y weihnachtsmann retrieve x y and x good man y good man neg same as x y gt x santa claus y weihnachtsmann x weihnachtsmann y santa claus The second notion of sameness if semantic equality Suppose you learn that weihnachtsmann is just the german name for santa claus thus weihnachtsmann and santa claus denote the same thing in the universe of discourse they are individual synonyms Let us add one more assertion that enforces that weihachtsmann and santa claus are synonyms same as weihnachtsmann santa claus gt OKAY individual synonyms santa claus gt weihnachtsmann santa claus In general for each individual there is a set of cardinality at least one of synonym individuals including the individual itself forming
87. the structure of the query head which may contain head projection operators and or lambda expressions let us discuss the struc ture of the complex query bodies Please note that some of the so far presented example queries already used complex query bodies A nRQL query body is inductively defined as being either a single query atom or a complex query body which is constructed from query bodies with the help of the following query body constructors As usual a prefix syntax is used e and is an n ary constructor which is used for the formulation of conjunctive queries The arguments of the and are called conjuncts The conjuncts are query bodies e union is an n ary constructor The arguments of the union are called disjuncts The constructor computes the union of the query answers of its disjuncts The disjuncts are query bodies e neg is a unary constructor the negation as failure NAF negation The argument is a query body e project to is a unary constructor This is the projection operator for query bodies not to be confused with with head projection operators The argument is a query body moreover the constructor also needs a so called projection list as argument So the first argument to this constructor is the projection list This list is a list of objects variables and or individuals Head projection operators are not permitted here The second argument is a query body Moreover the set of objects mentioned in the pro
88. the synonym equivalence class for that individual Racer Systems GmbH amp Co KG http www racer systems com 108 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL nRQL always removes redundant bindings from a query answer i e you will never get a query answer which contains x santa claus twice In order to detect redundant bindings nRQL uses by default name equality Thus nRQL considers two variable bindings as different if the individuals to which the variables are bound have different names retrieve x x good man gt x santa claus x weihnachtsmann However sometimes one wants to ensure that only semantically different bindings are delivered thus either x santa claus or x weihnachtsmann would be considered redundant This is possible as well see below In addition to the negated same as query atom nRQL offers a special role the so called nrql equal role which holds between two individuals if they are in the same synonym equivalence class So it becomes possible to query for pairs of individuals which have a different name but are in fact synonyms retrieve x y and neg same as x y x y nrql equal role gt x weihnachtsmann y santa claus x santa claus y weihnachtsmann Moreover nRQL can be switched into a mode in which variables are not bound to differ ent ABox individuals but to representative individuals from different synonym equ
89. this effect racer read file nrql user guide examples family racer gt OKAY instance doris mother OKAY v N retrieve x x some has child top x betty x doris x alice v N retrieve x x y has child x betty x alice v Now we have the situation for doris Since we have added the assertion instance doris mother it is the case that doris must have some child thus in every model of the KB there will be some successor of the has child role However this child is not explicitly present in the ABox Thus no binding can be found for y if x doris in x y has child It is possible to identifiy such individuals which have a certain logicall implied implicit role filler which is not explicitly modelled using the NAF negation neg and project to see below retrieve x and x mother neg project to x x y has child gt x doris Using a rule see below it even becomes possible to create some name for the logically implied child and add appropriate assertions for doris to the ABox as follows firerule and x mother neg project to x x y has child related x new ind child of x has child gt related doris child of doris has child Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 103 retrieve x y x y has child gt x doris y chil
90. type integer gt OKAY Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 119 instance i and equal width 10 equal length 20 OKAY Vv retrieve x lambda x y first x first y told value if exists width x told value if exists length x x and a width a length Px i lambda x y first x first y existing told values width x existing told values length x 200 v The function lambda application is performed by substituting the formal parameters x y to the actual arguments supplied by the two told value if exists projection operators These operators return lists of told values thus the first function is applied before is applied to yield the total size 6 1 3 1 MiniLisp MiniLisp is easy to understand and use for readers which have some Common Lisp CL experience MiniLisp only supports symbols strings numbers and lists Many of the built in operators are inherited from CL All RacerPro API functions can be called from within a lambda body RacerPro macros are treated as functions MiniLisp supports quote backquote comma and bq comma atsign This is very useful if nRQL sub queries with variable parts shall be executed from within a lambda body The lambda body is evaluated in an environment where current abox is bound to the query ABox and current tbox to the query TBox
91. vectors denote the structure of the tuple sets denoted by the query bodies The element positions of the tuples in these sets can be considered as typed e g the type of the tuples in and x woman y man is lt woman man gt In contrast the types of the tuples in and b woman a Racer Systems GmbH amp Co KG http www racer systems com 224 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL man is lt man woman gt So it is obvious that these two sets can never be in a subsumption subset relationship to one another Note that this is very similar to the covariance paradigm used in programming languages Let us give some practical examples gt full reset okay full reset racer read file home mi wessel nrql user guide examples family no signature racer OKAY prepare abox query nil x woman gt query 1 ready to run prepare abox query nil x grandmother gt query 2 ready to run query entails p query 2 query 1 gt t query entails p query 1 query 2 gt NIL prepare abox query nil and x woman x y has child gt query 5 ready to run prepare abox query nil z mother gt query 6 ready to run query entails p query 5 query 6 gt t query entails p query 6 query 5 gt NIL Please note that query 6 and query 5 are not equivalent Query 5 is stronger more specific since it requires the
92. width x told value if exists length x gt instance i size 200 retrieve x told value if exists size x x a size gt x i existing told values size x 200 Note that RacerPro does not even require a define primitive attribute size declaration here Note that we have modeled the size filler implicitly with the concept assertion instance i size 200 which has been added by the rule Sometimes one also wants to have CD objects e g size of i in order to be able to formulate further constraints with other CD objects A rule can also add constrained and constrained assertions Please consider the following rule which creates a new CD object called size of i representing i s size moreover we unnecessarily also add a constraints assertion on size of i firerule x and a width a length constrained x new ind size of x size constraints mew ind size of x lambda w 1 first w first 1 told value if exists width x told value if exists length x gt constrained i size of i size constraints size of i 200 retrieve x size x told value size x x a size gt x i size x size of i told values size x 200 Racer Systems GmbH amp Co KG http www racer systems com 158 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 1 6 7 Adding Pseudo Nominals with Rules Suppose
93. world assumption and the am biguities about what closing should actually mean in description logic inference systems usually there is no support for the closed world assumption However with the publish and subscribe interface of RacerPro users can achieve a similar effect Consider for instance a query for a book which does not have an author Because of the open world assump tion subscribing to a channel for and Book at most 0 has author does not make much sense Nevertheless the agent can subscribe to a channel for Book and a channel for at least 1 has author It can accumulate the results returned by RacerPro into two variables A and B respectively and in order to compute the set of books for which there does not exist an author it can consider the complement of B wrt A We see this strategy as an implementation of a local closed world LCW assumption However as time evolves authors for documents determined by the above mentioned query indeed might become known In others words the set B will probably be extended In this case the agent is responsible for implementing appropriate backtracking strategies of course The LCW example demonstrates that the RacerPro publish and subscribe interface is a very general mechanism which can also be used to solve other problems in knowledge representation Racer Systems GmbH amp Co KG http www racer systems com Chapter 6 The New RacerPro Query Language nRQL In
94. x and x woman x y has child y human x betty x alice v describe query last query 1 accurate processed retrieve x and y ano1 and human person x y ano1 has child x and woman mother abox smith family v DESCRIBE QUERY outputs by default always the internally rewritten query Note that nRQL has deduced that 333 x is in fact a mother not only a woman 333 You can also get the original query by 333 providing the optional nil argument describe query last nil gt query 1 Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 233 accurate processed retrieve x and x woman x y has child y human abox smith family retrieve x and x woman x all has descendant man x y has child y human gt NIL N describe query last gt query 2 accurate processed retrieve x and x and woman all has descendant man mother x y ano1 has child y ano1 and human man abox smith family Note that nRQL has deduced that y ano1 y must be a man and that x is a mother Since a realized query contains more syntactically explicit information the search for an swer tuples can eventually be more constrained and thus eliminate candidate bindings for variables which otherwise would be considered So
95. y betty x alice y charles retrieve x y meg x y has child gt x eve y eve x eve y doris x eve y charles x eve y betty x eve y alice x doris y eve x doris y doris x doris y charles x doris y betty x doris y alice x charles y eve Racer Systems GmbH amp Co KG http www racer systems com 136 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL x charles y doris x charles y charles x charles y betty x charles y alice x betty y charles x betty y betty x betty y alice x alice y eve x alice y doris x alice y alice Note that 4 21 25 5 5 and we have 5 individuals in that KB Suppose we are now looking for people without known children So we try retrieve x neg x y has child gt x eve x doris x charles x betty x alice However we already know that only betty and alice have children in the ABox retrieve x x y has child gt x betty x alice In fact we were looking for the answer x charles x eve x doris Didn t we just say that neg lt atom gt always returns the complement set of lt atom gt So what went wrong In fact internally the correct complement set is constructed However these two dimensional sets are no longer complementary to
96. 1 The Query Processing ModiofnRQL 203 6 2 2 The Life Cycle of a Query 0 0 2 022 82s 205 6 2 3 The Life Cycle of a Rule 207 6 2 4 How to Implement Your Own Rule Application Strategy 208 6 2 5 Configuring the Degree of nRQL Completeness 212 6 2 6 Automatic Deadlock Prevention LL 217 6 2 7 Reasoning with Queries oaaao 220 6 2 8 The Query Repository The QBox 4 227 62 9 The Query Realizer oe sa sa 085 a daaa E ee a eS 232 6 2 10 The nRQL Persistency Facility aoaaa 234 7 Outlook 237 A Another Family Knowledge Base 239 B A Knowledge Base with Concrete Domains 241 C SWRL Example Ontology 247 D LUBM benchmark 251 Index 261 Racer Systems GmbH amp Co KG http www racer systems com Chapter 1 Introduction RacerPro stands for Renamed ABox and Concept Expression Reasoner Professional As the name suggests the origins of RacerPro are within the area of description logics Since description logics provide the foundation of international approaches to standardize ontology languages in the context of the so called semantic web RacerPro can also be used as a system for managing semantic web ontologies based on OWL e g it can be used as a reasoning engine for ontology editors such as Prot g However RacerPro can also be seen as a semantic web information repository with optimized retrieval engine because it can handle large sets of data de
97. 2001 XMLSchema int gt lt rdf type rdf resource http www w3 org 2002 07 owl AnnotationProperty gt lt owl DatatypeProperty gt lt class rdf ID test gt lt rdfs comment rdf datatype http www w3 0rg 2001 XMLSchema string gt Comment for individual test lt rdfs comment gt lt object property rdf resource test gt lt datatype property rdf datatype http www w3 0rg 2001 XMLSchema int gt 123 lt datatype property gt lt annotation object property rdf resource test gt lt annotation datatype property rdf datatype http www w3 org 2001 XMLSchema int gt 456 lt annotation datatype property gt lt class gt lt rdf RDF gt Thus here we have e an ontology which is annotated with the string Demo Ontology which is the filler of the RDF S predefined rdfs comment annotation datatype property of type XSD string e a class class which is rdfs comment annotated with This is a test class e an ordinary object property object property e an ordinary datatype property datatype property of type XSD int e an annotation object property annotation object property e an annotation datatype property annotation datatype property of type int e and an individual named test being an instance of class which has itself as a filler of object property the integer 123 as a filler of datatype property Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LA
98. ACERPRO QUERY LANGUAGE NRQL x y z u and x y has descendant x z has descendant y u has descendant z u has descendant neg same as y z neg y z has descendant neg y z has ancestor neg y z has synonym gt NIL So there are no non trivial diamonds in the taxonomy We can add one more node to the taxonomy which produces some diamonds in order to avoid getting all the synonym matches as well we constrain the variable variables to taxonomy node individuals only tbox retrieve x y z u and x y has descendant x z has descendant y u has descendant z u has descendant neg same as y z neg y z has descendant neg y z has ancestor neg y z has synonym x taxonomy node y taxonomy node z taxonomy node u taxonomy node gt x top y man z woman u bottom x top y man z woman with children u bottom x top y woman with children z man u bottom x top y woman z man u bottom Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 167 6 1 8 Hybrid Representations with the Substrate Representation Layer The so called substrate representation layer is used to associate a RacerPro ABox with an additional representation layer This additional representation layer is called a substrate The coupling of a RacerPro ABox with a substrate re
99. AN 2 betty rdfs comment P family MALE gt farnily MAN Y family PERSON v familyPARENT family FATHER family MOTHER 2 gt family WOMAN family HAS GENDER va MEJ ay Name Prefix Conditions O AUNT family family WOMAN n 3 family HAS SIBLING family PARENT a a AllDifferent owl E owlThing E P AllvaluesFromRestriction owl E ow Restriction jo Alt rdfs E rdfs Container AnnotationProperty owl c rdfProperty z 5 AnonymousClass owl E OWL CLASS BROTHER family family MAN n 3 family HAS SIBLING family PERSON B Bag rdfs E rdfs Container baj cc Classification Results Subclasses of owl Thing Figure 2 19 Screenshot of Prot g displaying inferred information about UNCLE If you press the I button in Prot g the inferred types are computed for the individuals aka instances or nominals in the ontology see also Figure 2 19 Although Prot g can be used to display and edit A boxes let us return to RICE for a moment to examine the result of querying an A box In our example we assume that the RICE window is still open Select the concept PERSON in the concept window The instances of PERSON are displayed in the upper right instance window 2 5 3 Using Prot g and RacerPorter in Combination Let us assume RacerPro is started Prot g is
100. ANCE BROWSER INDIVIDUAL EDITOR Pio BT french italian owl Thing I gt john instance of owl T l Thing 1 1 SENI NAM ME x Name SameAs DifferentFrom IC D swrl Imp 2 O swri Variable 1 D john john lol swrl Builtin 78 P French e Italian Wine drinker a r French or Italian 1 be E TE ow No wine drinker owl Thing X French or Italian i No wine drinker coni CIA Lelke Figure 2 36 For john we have indefinite information John is an instance of French or Italian In RacerPro 1 9 for a fixed variable binding to individual in an A box the predicates in the body must be satisfied in all models This has important consequences Let us assume we have class names French Italian Wine drinker Not wine drinker and French or italian For the latter two classes necessary and sufficient conditions are spec ified see Figures 2 34 and 2 35 The individuals john is an instance of Not wine drinker and French or Italian Racer Systems GmbH amp Co KG http www racer systems com 2 6 SWRL SEMANTIC WEB RULE LANGUAGE 53 18009 french italian Prot g 3 1 1 file Applications Protege_3 1 french italian ppri OWL Files owl or rdf Die e e a es gt E 0 E RPEEEI N J prot g Cj OWLClasses P Properties Forms lt I gt Individuals am
101. CERPRO QUERY LANGUAGE NRQL betty doris has child alice charles has child alice betty has child 6 2 5 Configuring the Degree of nRQL Completeness So far we have only used the complete modi of nRQL In these modi nRQL uses the basic ABox querying functions of RacerPro in order to achieve completeness However on very big or pathological KBs this mode might fail to scale In such cases you might consider trying the incomplete modi of nRQL For ABoxes KBs which have a simple structure for example ABoxes created from RDF S documents the incomplete modi will be complete Many ABoxes simply contain data those ABoxes are basically sets of ground terms and thus contain no disjunctions no role qualifications etc Bulk data from relational databases would qualify for the incomplete modi Note that in RDMS usually the Closed World As sumption CWA is made in contrast to RacerPro which uses the Open World Assumption OWA However note that nRQL offers negation as failure and thus some form of CWA negation is available In order to query such a potentially huge but simply structured ABox nRQL can be config ured in such a way that it uses only the syntactic told information from the ABox for query answering and exploits also information from the taxonomy to achieve better completeness Thus query answering can be implemented using plain relational lookup techniques and only shallow reasoning is required As mentio
102. Concept Query to retrieve the instances of the selected concept Note that the Concept field of the state display always shows the current Concept Adapt the Tree Depth accordingly Note that Tree Depth is ignored if Graph display mode is selected If you push Focus while in Tree display mode then only the subconcepts of the current Concept will be shown In the Individuals tab select an individual press Direct Types or All Types then go to the Taxonomy tab It will highlight the types of the selected individual note the Individual field in the state display Select the Network tab afterwards in order to show the ABox structure focusing on the selected individual Adapt the Tree Depth accordingly Then push Focus You can always change the focus individual by simply selecting it The focus individual is shown in the Individual field at the top To display the complete ABox structure simply push the Reset button Note that Tree Depth is ignored if Graph display mode is selected You can also focus on a subset of the ABox edges simply select the roles you are interested in from the Roles tab then check out Selected Roles Roles Tab in the Network tab Then only the edges labeled with selected roles are shown Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO 29 Note that the Roles tab is the only tab which allows multiple selection of items The last selected
103. Consequence nRQL rules have an antecedent or precondition and a consequence or postcondition An antecedent is just an ordinary nRQL query body The consequence is a set of so called generalized A Boz assertions A generalized ABox assertion can also reference variables which are bound in the rule antecedent query body A rule consequence can contain generalized concept assertions generalized role assertions generalized constrained generalized constraint assertions generalized same as asser tions Moreover assertions can also be removed 6 1 6 2 A Simple nRQL Rule The following simple nRQL rule promotes woman which are not yet mothers to mothers full reset okay full reset v racer read file nrql user guide examples family racer OKAY v firerule and x woman neg x mother instance x mother gt instance doris mother instance eve mother The antecedence is the query body and x woman neg x mother and the con sequence is a single generalize instance assertion instance x mother As shown the variables in the generalized assertions in a rule consequence are instantiated with the bind ings satisfying the rule antecedence i e the bindings which have been computed by the query body in the antecedence The constructed assertions are returned as the result of the rule application We can easily verify with a query that eve and doris have become mothers as well
104. D Person gt lt owl ObjectProperty rdf ID hasChild gt lt owl ObjectProperty rdf ID hasSibling gt lt swrl Imp rdf ID Rule 1 gt lt swrl body gt lt swrl AtomList gt lt rdf first gt lt swrl IndividualPropertyAtom gt lt swrl argument2 gt lt swrl Variable rdf ID y gt lt swrl argument2 gt lt swrl propertyPredicate rdf resource hasChild gt lt swrl argument1 gt lt swrl Variable rdf ID x gt lt swrl argument1 gt lt swrl IndividualPropertyAtom gt Racer Systems GmbH amp Co KG http www racer systems com 248 APPENDIX C SWRL EXAMPLE ONTOLOGY lt rdf first gt lt rdf rest gt lt swrl AtomList gt lt rdf rest gt lt swrl AtomList gt lt rdf first gt lt swrl DifferentIndividualsAtom gt lt swrl argument2 gt lt swrl Variable rdf ID z gt lt swrl argument2 gt lt swrl argumenti rdf resource y gt lt swrl DifferentIndividualsAtom gt lt rdf first gt lt rdf rest rdf resource http www w3 org 1999 02 22 rdf syntax ns nil gt lt swrl AtomList gt lt rdf rest gt lt rdf first gt lt swrl IndividualPropertyAtom gt lt swrl propertyPredicate rdf resource hasChild gt lt swrl argument1 rdf resource x gt lt swrl argument2 rdf resource z gt lt swrl IndividualPropertyAtom gt lt rdf first gt lt swrl AtomList gt lt rdf rest gt lt swrl AtomList gt lt swrl body gt lt swrl head gt lt swrl AtomList gt
105. Disconnect Profiles Refresh ClearLog FullReset New Open Load RacerPro Log hasUncle x 2 lt A hasMother x y hasBrother y 2 P il done 6 gt temp temp 1 owL 7 CSWRL FORWARD CHAINING Execute http www owl ontologies com unnamed owl1 Rule 11 RELATED http waw owL ontoLlogies com unnamed owlL MARY http waw owL ontoLogies com unnamed owL JOHN http ww owl ontologies com unnamed owl hasUncle 7 gt 0KAY O Figure 2 31 The uncle relation is established In Figure 2 31 we see that RacerPro fired the rule accordingly The network tab of Racer Porter allows us to analyze the effect graphically see Figure 2 32 With RacerPorter rules can be inspected rerun etc see Figure 2 33 Racer Systems GmbH amp Co KG http www racer systems com 2 6 SWRL SEMANTIC WEB RULE LANGUAGE 49 E 7 090 RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox temp 1 0wl ABox temp 1 owl Concept Role Individual MARY Query Rule v Simplify Names JOHN MARY has Unc ChasMother SUE sBrother Tree Depth 1 E O Tree Graph Transitive CU Selected Roles Roles Tab 0 Horizontally O Vertically Focus Reset C Clear Log 3 Direct Types 3 E All Types Consistent 3 E Realize 3 RacerPro Log 7 CSWRL FORWARD CHAINING Execu
106. E NRQL QUERY PROCESSING ENGINE 213 as instance i and a b is found then also instance i a and instance i b will be added Please note that these concept assertions are only virtually added the original ABox is not modified Thus retrieve x x a and retrieve x x b will succeed Regarding the relational structure of the ABox which is spawned by the role assertions nRQL is only complete in this mode if there are no at most number restrictions present in the ABox and if there are no features used Thus nRQL will be complete up to the description logic ACCHZ p concerning the relational structure of the ABox even in this incomplete mode This means the effects of role hierarchies transitively closed roles and inverse roles are completely captured If you consider to query a plain relational structure and you don t need a TBox with labeled nodes for example a big graph representing a public transportation network with a few 10 000 nodes or the like then this is the nRQL setting you will need and which is complete enough for your application You could also use the data substrate In order to enable this mode use set nrql mode 0 Told information querying plus taxonomic information for atomic concepts con cept names Mode 1 In addition to the virtual concept membership assertions already added in Mode 0 this mode considers the concept assertions of the form instance i C for atomic
107. ESCENDANT HAS DESCENDANT Point pushed A Figure 2 6 The Console tab of RacerPorter used for executing queries Racer Systems GmbH amp Co KG http www racer systems com 2 1 SAMPLE SESSION WITH RACERPORTER 17 e ee RacerPorter i Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox FAMILY ABox SMITH FAMILY Concept Role Individual Query Rule CC Simplify Names CHARLES ALICE HAS CHILD CHAS CHILDJ BETTY EVE BETTY HAS CHILD CHAS CHILD DORIS CHARLES BETTY _ GHAS SIBLING DORIS __ EVE _ GHAS SISTER HAS SIBLING EVE gt DORIS _ GHAS SISTER HAS SIBLING Tree Depth 2 HH tree O Graph Transitive Selected Roles Roles Tab Horizontally Vertically Focus Reset ClearLog DirectTypes All Types Consistent Realize RacerPro Log 6 CALL TRANSITIVE ROLES 6 gt CCINV HAS DESCENDANT HAS DESCENDANT 7 CALL INDIVIDUALS 7 gt CEVE DORIS CHARLES BETTY ALICE Figure 2 7 The Network tab shows the explicit relations for A box individuals RacerPorter allows you to inspect data descriptions in an A box Just select the Network tab see Figure 2 7 Adjust the display options to Tree and select Tree Depth 3 as shown in Figure 2 7 Racer Systems GmbH amp Co KG http www racer systems com 18 CHAPTER 2 USING RACERPRO 2 2 The RacerPro Server The RacerP
108. ESSING ENGINE 211 gt NIL execute or reexecute rule rule 1 gt rule 1 running execute rule rule 1 inot found get next set of rule consequences rule 1 gt instance betty mother related betty child of betty has child applicable rules gt NIL choose current set of rule consequences rule 1 gt instance betty mother related betty child of betty has child get next set of rule consequences rule 1 gt instance alice mother related alice child of alice has child add chosen sets of rule consequences rule 1 NRQL WARNING DENIED DUE TO DEADLOCK PREVENTION THE FOLLOWING RULES WILL NOT TERMINATE AUTOMATICALLY SINCE THEY HAVE BEEN STARTED IN LAZY INCREMENTAL MODE rule 1 gt denied due to deadlock prevention get next n remaining sets of rule consequences rule 1 gt instance eve mother related eve child of eve has child instance doris mother related doris child of doris has child get next n remaining sets of rule consequences rule 1 gt warning kb has changed get next n remaining sets of rule consequences rule 1 gt NIL all role assertions gt betty child of betty has child alice child of alice has child eve doris has sister doris eve has sister charles betty has sibling betty eve has child Racer Systems GmbH amp Co KG http www racer systems com 212 CHAPTER 6 THE NEW RA
109. F in constraint query atom 128 NAF in query with individuals 128 NAF in SAME AS query atom 128 NAF query 124 negation as failure 124 nominals 70 142 nominals in queries 142 nRQL API 172 nRQL EBNF 164 nRQL engine 172 nRQL grammar 164 nRQL mode 179 nRQL mode 0 179 nRQL mode 1 180 nRQL mode 2 181 nRQL mode 3 181 Racer Systems GmbH amp Co KG http www racer systems com 264 INDEX nRQL mode 4 181 nRQL mode 5 181 nRQL mode 6 181 nRQL modes 172 181 nRQL persistency 201 nRQL syntax 164 number restriction 57 object 92 open world assumption 72 optimization strategies 77 options 22 OWL 74 OWL annotation property filler query 115 OWL annotation property value query 115 OWL constraint checking 106 OWL constraint query atom 106 OWL datatype property filler query 115 OWL datatype property value query 115 OWL instances query 95 OWL mirror 154 158 OWL object property filler query 100 OWL query 95 100 106 115 158 persistency 78 predicate 153 predicate query 153 preferences 25 primitive concept 60 projection 130 projection operator 130 pseudo nominal 142 publish subscribe mechanism 79 QBox 195 qualitative spatial reasoning 161 query API 172 query atom 92 query body 93 query cache 195 query conjunct 119 query consistency 188 query disjunct 120 query engine 172 query entailment 190 query head 93 110 query incon
110. However it is not since x and y are different individuals The disjuncts of the union have thus arity 2 describe query last gt query 11 accurate processed retrieve x or and x man top y ano1 and top x y ano1 woman abox smith family nRQL has rewritten this body into union and x man top y ano1 and top x y ano1 woman the variable y has been replaced with y ano1 since y is not mentioned in the query head Since we are thus only interested in the bindings of x it becomes clear that this query body is in fact equivalent to union x body x top or x top Another pitfall is to think that the query retrieve x union x C x not C is equivalent to retrieve x x or C not C However it is not since the concept or C not C is equivalent to top and thus all individuals will be returned However the union query returns only those individuals for which either C or not C can be proven Due to the Open World Semantics OWA there can be individuals which are neither instances of C nor of its complement not C but every individual is of course an instance of top Racer Systems GmbH amp Co KG http www racer systems com 130 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL full reset gt okay full reset instance i c gt OKAY instance j not c gt OKAY instance k top gt OKAY retr
111. L API provides the function get next tuple for this purpose Applications can also check for the availability of a next tuple by calling next tuple available p process tuple at a time gt o0kay processing tuple at a time retrieve x x woman gt query 3 running get next tuple query 3 gt x alice get next tuple last gt x betty next tuple available p last gt t get next tuple last gt x eve next tuple available p last gt t get next tuple last gt x doris next tuple available p last gt NIL get next tuple last gt exhausted get answer last gt x alice x betty x eve x doris describe query last query 3 accurate processed retrieve x x woman abox smith family Vow In the incremental mode a call to retrieve does not return the query answer but instead returns a so called query identifier ID which can also be understood as a kind of iterator This ID is used as argument to API functions such as get next tuple Moreover the identifier last always refers to the last submitted query or rule In order to differentiate queries from rules also last query and last rule can be used instead of last Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 205 6 2 1 2 1 Concurrent Incremental Query Processing The nRQL eng
112. NGUAGE 117 the string Comment for individual test as a filler of the annotation datatype property rdf comment itself as a filler of annotation object property and the integer 456 as as filler of the annotation datatype property The following queries demonstrate how the fillers of these different properties can be re trieved with nRQL the queries should speak for themselves retrieve x x class x http www owl ontologies com Ontology1159352693 owl test Vv retrieve x y x y object property x http www owl ontologies com Ontology1159352693 owl test y http www owl ontologies com Ontology1159352693 owl test Vw retrieve x y x y annotation object property x http www owl ontologies com Ontology1159352693 owl test y http www owl ontologies com Ontology1159352693 owl test Vv retrieve x x some object property class x http www owl ontologies com Ontology1159352693 owl test Vow retrieve x x some annotation object property class NIL Vow Note no reasoning for annotation object properties retrieve x datatype fillers datatype property x x a datatype property x http www owl ontologies com Ontology1159352693 owl test datatype fillers http www owl ontologies com Ontology1159352693 owl datatype property x 123 Vv retrieve x da
113. Netw Conc Roles indivi Asser Queries Rules TBox Ihttp dl kr org dig kb 471611 ABox Ihttp dl kr org dig kb 471611 Concept Role Individual Query Rule l Simplify Names CiGreenPer a ISundried Si cedri gt lt Figure 2 23 RacerPorter used to inspect an ontology edited and verified with Prot g Prot g uses the DIG interface to connect to RacerPro Therefore one can use RacerPorter to inspect details about the pizza knowledge base in Figure 2 23 a graph of the taxonomy of the pizza ontology is presented With RacerPorter for instance queries can be answered Figure 2 24 shows the result of a queries for countries used in pizza owl Queries can be inspected also using the tab Queries in RacerPorter see Figure 2 23 Racer Systems GmbH amp Co KG http www racer systems com 42 CHAPTER 2 USING RACERPRO 6 00e RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox Ihttpy di kr org dig kb 471611 ABox Ihttpy di kr org dig kb 471611 Concept Role Individual Query Rule Simplify Names Disconnect Profiles Refresh ClearLog FullReset New Open Load RacerPro Listener all individuals CiItaly Germany IFrancel lEnglandl lAmerical Cretrieve x x Country CCC X IFrancel CC X Germany X lAmerical X lItalyl X lEngland
114. Nominals or Concrete Domains The language OWL provides for means to address individuals in concepts As a example one could represent the concept human with the following axioms define primitive role ancestor of transitive t inverse has descendant define primitive role has descendant equivalent human some ancestor of one of adam instance john human related kain john has descendant The concept forming operator one of takes an individual and construct a concept whose extension contains just the semantic object to which the individual adam is mapped Thus the extension of one of adam is a singleton set Individuals in concepts are also known a nominals Asking for the ancestors of john yields kain and adam although the latter is not explicitly stated in the A box Currently RacerPro does not support nominals in full generality Only axioms of the form equivalent one of i c equivalent one of i some r one of j are supported These axioms are very important for practical purposes and they directly correspond to the following A box assertions Racer Systems GmbH amp Co KG http www racer systems com 4 2 NOMINALS OR CONCRETE DOMAINS 71 instance i c related i j r In many practically important cases nominals are not required in concepts terms The same effect can be achieve using concrete domain values For instance one might think of using nominals for representing colors of simp
115. ODUCTION e Find implicit subclass relationships induced by the declaration in the ontology e Find synonyms for resources either classes or instance names e Since extensional information from OWL documents OWL instances and their inter relationships needs to be queried for client applications an OWL QL query processing system is available as an open source project for RacerPro e HTTP client for retrieving imported resources from the web Multiple resources can be imported into one ontology e Incremental query answering for information retrieval tasks retrieve the next n re sults of a query In addition RacerPro supports the adaptive use of computational resource Answers which require few computational resources are delivered first and user applications can decide whether computing all answers is worth the effort Future extensions for OWL e g OWL E are already supported by RacerPro if the sys tem is seen as a description logic system RacerPro already supports qualified cardinality restrictions as an extension to OWL DL 1 1 2 RacerPro as a Description Logic System RacerPro is a knowledge representation system that implements a highly optimized tableau calculus for a very expressive description logic It offers reasoning services for multiple T boxes and for multiple A boxes as well The system implements the description logic ALCOHT p also known as SHTQ see 9 This is the basic logic ALC augmented with qualifying
116. ROCESSING ENGINE 231 show qbox for abox smith family t 333 QBOX FOR racer dummy substrate FOR ABOX smith family master top query seen and x ano1 woman x anol y has child y human js ___ and x anol mother y x ano1 has parent y man i330 ___master bottom query sir x top NE ___ a human a3 ___ b woman vi ___ a mother ia ___master bottom query gt see output on stdout Racer Systems GmbH amp Co KG http www racer systems com 232 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 2 9 The Query Realizer Query realization can be called a semantic query optimization technique which will enhance the amount of information that is available for the heuristic cost based query optimizer Moreover the added conjuncts can also enhance the overall query answering speed since the search process computing the answer tuples is somewhat better informed if more information regarding the query is made explicit by means of inference However the service is still experimental The query realization process is similar to an ABox realization process logically implied conjuncts are added The realized query is semantically equivalent to the original query Let us consider the following example session full reset gt okay full reset N racer read file OKAY v J enable query realization okay query realization enabled Vv N retrieve
117. RY LANGUAGE NRQL in abox geo example 333 Create an RCC5 substrate which is associated with the ABox 333 geo example in rcc box geo example rcc5 333 Create some RCC substrate nodes 333 Note that data node can be used 33 as well rcc instance europe rcc instance germany country germany and country string has name Germany all has language german some has language language inhabitants 82600000 rcc instance france country france and country string has name France all has language french some has language language inhabitants 60656178 rcc instance hamburg city hamburg and city some in country germany string has name Hamburg rcc instance paris city paris and city some in country france string has name Paris 333 Create some RCC substrate edges 333 Europe contains Germany rcc related europe germany ppi rcc related europe france ppi Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 189 rcc related germany hamburg ppi rcc related france paris ppi rcc related france germany dr In this example 5 RCC nodes are created europe germany france paris hamburg and it is stated that europe contains france and germany which do not overlap Moreover paris is contained in france and hamburg in germany These are the only spatial relation ships given For all nodes but europe
118. RacerPro User s Guide Version 1 9 2 Racer Systems GmbH amp Co KG http www racer systems com October 18 2007 Contents 1 Introduction LI 1 2 1 3 1 4 2 1 2 2 2 3 2A Views ai RacerFro e co ece e E ee ke eR ce oe Be ed 1 1 1 RacerPro as a Semantic Web Reasoning System and Information PRREPOSICOnE lt s oh ae a SG A Da ee Be Re e ee ow ee A 1 1 2 RacerPro as a Description Logic System 1 1 3 RacerPro as a Combination of Description Logic and Specific Rela tional Algebras co hee e e RE eS Application Areas and Usage Scenarios n oo sooo e Racer Editions Installation Licenses and System Requirements bot Ione coea e e e oe we E e eo PE oe de amp aoe Woe Installation gt Gava e E BAO e a AE SA e es 133 Licenses pos ne Song RARE ee eae Re Ae Ge a ee 1 3 4 System Requirements ss so sa cera aaan 020052 eee eee Ackowledment es s s ee 44 88 ee Rbk RR RY ke he ees Using RacerPro Sample Session with RacerPorter 2 0 eee eee eee The RacerPro Servet oo ccie data aea a Pee he Rew Es 2 2 1 The File Interface c soo soo 6 82544 868544 be ba ee ees 20 TOPAPIG La ton bop we te ea doa BOS ai n 2 2 3 Web Service Interface oc sor eg e Ok e Re ee tes 2 2 4 HTTP Interface DIG Interface 2 0 2 2 2 5 Options for the RacerPro Server How to Send Bug Reports o se sads c rona etia mi e a E a a Racer F OIER oe ca eae Se ee EOR IA e ee eR a a a e E
119. TENT gt inconsistent retrieve nil and x grandmother x not woman NRQL WARNING query 7 IS INCONSISTENT gt inconsistent concept satisfiable and woman man gt NIL prepare abox query nil betty man gt query 8 ready to run query consistent p last gt t 6 2 7 2 Reporting Inconsistent Rules The described reasoning services are also provided for rule In this case also the consequence of the rule is taken into account full reset okay full reset v report inconsistent queries and rules v okay reporting inconsistent queries and rules prepare abox rule and x c y d instance x d related x y r instance y all inv r not d NRQL WARNING rule 1 IS INCONSISTENT gt rule 1 ready to run prepare abox rule and x c x not c instance x d NRQL WARNING rule 2 IS INCONSISTENT gt rule 2 ready to run prepare abox rule and x c x d instance x not and c d NRQL WARNING rule 3 IS INCONSISTENT gt rule 3 ready to run Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 223 6 2 7 3 Checking for Query Entailment The function query entails p checks whether one query is more specific than another one This is also known as query subsumption The function query equivalent p checks whether both queries mutually entail subsume
120. UBM BENCHMARK 333 Query 11 benchmark universities departments 11 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl ResearchGroup x http www University0 edul Inttp www lehigh edu X7Ezhp2 2004 0401 univ bench owl subOrganizationOf x 333 Query 12 benchmark universities departments 12 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Chair y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Department x y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl memberOf y http www University0 edul http www lehigh edu 7Ezhp2 2004 0401 univ bench owl sub0rganizationOf x y 333 Query 13 benchmark universities departments 13 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Person Inttp www University0 edul CE http www lehigh edu 7Ezhp2 2004 0401 univ bench owl hasAlumnus x 333 Query 14 benchmark universities departments 14 x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl UndergraduateStudent x defun simple lubm universities departments benchmark universities departments 1 x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Chair x benchmark universities departments 2 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Chair y Ihttp www lehigh edu 7Ezhp2 2004 0401 univ bench owl Department x y http www lehigh
121. WL KB thus test means http www owl ontologies com Ontology1159352693 owl test However we al ready mentioned that nodes in the substrate must be prefixed with an additional So to access the node http www owl ontologies com Ontology1159352693 owl test from a substrate query atom please either remember to add a in front of the qualified name or use the amp prefix instead the following queries demonstrate this retrieve x same as x http www owl ontologies com Ontology1159352693 owl test gt ERROR NRQL ERROR Parser Error Unknown expression same as x http www owl ontologies com Ontology1159352693 owl test found retrieve x same as x http www owl ontologies com Ontology1159352693 owl test gt x http www owl ontologies com Ontology1159352693 owl test retrieve x same as x test gt ERROR NRQL ERROR Parser Error Unknown expression same as x http www owl ontologies com Ontology1159352693 owl test found retrieve x same as x amp test gt x http www owl ontologies com Ontology1159352693 0wl test Some more queries should be sufficient to demonstrate the utility of the mirror substrate retrieve x Pry and x owl object x y owl datatype property relationship y owl literal predicate numberp predicate gt 100 predicate lt 200 gt x http www owl onto
122. about services descriptions e Medicine Bioinformatics e g represent and manage biomedical ontologies e Natural Language Processing and Knowledge Based Vision e g exploit and combine multiple clues coming from low level vision processes in a semantically meaningful way e Process Engineering e g represent service descriptions e Knowledge Engineering e g represent ontologies e Software Engineering e g represent the semantics of UML class diagrams 1 3 Racer Editions Installation Licenses and System Re quirements 1 3 1 Editions Racer Systems provides several different editions of the Racer technology RacerPro RacerPro is a server for description logic or OWL inference services With Rac erPro you can implement industrial strength projects as well as doing research on knowledge basis and develop complex applications If you do not have a valid license you are allowed to use RacerPro but some restrictions apply RacerPorter The Porter to RacerPro aka RacerPorter is the graphical user client for RacerPro RacerPorter uses the TCP IP network interface to connect to one or more RacerPro servers and helps you manage them You can load knowledge bases switch between different taxonomies inspect your instances visualize T Boxes and A Boxes manipulate the server and much more RacerPorter is already included in the installer versions of RacerPro for Windows and Mac OS X and separately available for Linux systems
123. age See below for examples The nRQL language must be distinguished from the nRQL query processing engine which is an internal part of the RacerPro server The main features of this query processing answering engine are A cost based heuristic optimizer The optimizer reorders the query atoms in a con junction in order to find a performant query evaluation plan In order to estimate the costs caused by a query atom ABox statistics cardinality information are exploited The engine is also a rule engine A rule has an antecedence precondition left hand side which is a nRQL query The consequence postcondition right hand side con tains so called generalized ABox assertions ABox assertions referencing variables and or individuals Queries and rules are maintained as objects within the engine Query objects under stand a complex protocol i e they can be asked to return their stored answers be deleted reexecuted etc A query rule is identified with an Id The engine offers a full life cycle management for queries and rules The engine supports defined queries A defined query is a query which has been given a name not to be confused with the Id of a query rule although Id and name can be identically The named query can be used to refer to that query within other queries similar to a macro mechanism Definitions must be acyclic The engine supports multi processing of queries which means that more than one quer
124. ait if it is about to be executed and there is no thread available in the pool It must thus wait until a thread is released by another currently running query and put back into the pool e We already mentioned the substrate representation layer The engine offers an exten sive API for maintaining this layer There are different kinds classes of representation substrates available tailored for the representation of different aspects According to the substrate kind different substrate query atoms apply e Reasoning on queries This functionality is currently still experimental and incom plete nRQL offers incomplete query satisfiability and query containment sub sumption checks nRQL can be advised to maintain and exploit a so called query repository QBox which is a DAG structured query cache nRQL uses the query containment check to Racer Systems GmbH amp Co KG http www racer systems com 92 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL maintain and compute this DAG This can be called an semantic optimization unlike the heuristic optimizer which uses only syntactic information A further semantic but still experimental optimization technique which relies on query reasoning called query realization is also implemented This Chapter is structured as follows e The first section introduces the nRQL query language in a tutorial like style e The second section describes the nRQL engine and its
125. al objects to one another They are closely related to natural language prepositions such as contains overlaps adjacent etc 6 1 8 3 1 The Purpose of the RCC Substrate The RCC substrate is a special kind of data substrate which offers representation and querying facilities for so called RCC networks The edges of an RCC network are labeled with RCC relationships denoting the relative qualitative spatial arrangement of two objects The RCC substrate offers the following sets of RCC relationships e The set of RCC5 relationships distinguishes 5 relationships DR discrete from mean ing that the two objects neither overlap nor are adjacent PO partial overlap the two objects overlap partially EQ equal the two objects are congruent PP proper part meaning the first object is contained within the second one and its inverse PPI proper part inverse denoting containment e The set of RCCS relationships is a finer version of RCC5 It is finer than RCC5 because it also considers the relationships of the borders of the objects to refine the qualitative spatial description RCC8 includes the RCC5 relationships PO and EQ RCC8 distinguishes in the DR case whether two non overlapping objects are adjacent externally connected EC or whether they are disconnected DC Thus the RCC5 relation DR is refined into the EC and DC case Moreover the PP and PPI relations are Racer Systems GmbH amp Co KG http www racer systems c
126. ally want to refer to the defined query please use the substitute keyword as illustrated above 6 1 5 2 Using Defined Queries in Query Definitions Of course it is possible to use defined queries in query definitions However cyclic definitions are prohibited defquery mother of male child m and substitute mother of m c c man gt mother of male child retrieve x substitute mother of male child x gt x alice The alternative but ambiguous syntax can be used as well 6 1 5 3 The API for Defined Queries There are various nRQL API functions for accessing and manipulating the defined queries Please refer to the Reference Manual 6 1 5 4 Defined Queries Belong to the TBox Please note that the definitions are local to the current tbox Thus if the current TBox changes the definitions are changing too If you want to put a definition into a TBox other than the current one you can supply an optional keyword argument tbox to defquery and related API functions and macros Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 151 Vv full reset okay full reset in tbox a gt a in abox a gt a defquery test x x a gt test instance a a gt OKAY retrieve x x test gt x a describe all definitions gt defquery test x x a in tbox b b in abox b b
127. amples people pets owl gt nrql user guide examples people pets owl Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 159 So let us ask which persons have no pet retrieve x and x person neg project to x x y has_pet gt x http cohse semanticweb org ontologies people Kevin Let us create cat pets for those people with a rule firerule and x person neg project to x x y has_pet related x new ind cat of x has_pet instance new ind cat of x cat gt related http cohse semanticweb org ontologies people Kevin http cohse semanticweb org ontologies people cat of Kevin http cohse semanticweb org ontologies people has_pet instance http cohse semanticweb org ontologies people cat of Kevin http cohse semanticweb org ontologies people cat Let us verify that kevin has a cat now retrieve pet of kevin direct types pet of kevin Kevin pet of kevin has_pet dont show lambdas p t gt pet of kevin http cohse semanticweb org ontologies people cat of Kevin http cohse semanticweb org ontologies people cat http cohse semanticweb org ontologies people pet 6 1 6 9 The API of the Rule Engine There is an extensive rule API Please consult the reference manual Moreover rules can be used in a add all consequences at once or in a incremental add one
128. an example the family knowledge base might be loaded into the server at startup RacerPro u init family racer Or under windows type RacerPro u init family racer Then either double click the jar file icon or type java jar rice jar into a shell Connect RICE to RacerPro by selecting Connect from the Tools menu In addition to the default T box named default always provided by RacerPro the T box FAMILY is displayed in the left upper window which displays the parent children relationship between concept names of different T boxes An example screenshot is shown in Figure 2 10 Users can interactively unfold and fold the tree display Racer Systems GmbH amp Co KG http www racer systems com 30 CHAPTER 2 USING RACERPRO amp RACER Interactive CLIENT Environment RICE File Edit Tools Help Concepts Type a Concept Name to show exact matching only ABoxes B S FAMILY 1 FEMALE EA HUMAN Ey PERSON EQ MAN i ES BROTHER Som UNCLE em FATHER E PARENT i ie FATHER i Ele MOTHER i Son GRANDMOTHER Ey WOMAN E MOTHER 2 SISTER Statements and or concept definitions concept instances or man woman PreClassify RACER Replies DORIS EVE BETTY CHARLES ALICE Figure 2 10 Screenshot of RICE Using the middle subwindow statements declarations and queries can be typed and sub mitted to RacerPro The answers are printed into the lower window A query concerning the
129. ano2 constraint age age lt abox smith family Instead of project to also project and pi can be used as keywords Please note that neg in combination with project to allows to formulate some kind of closed world closed domain universal quantification as just demonstrated Since there ex ists no know person that is older than x we know that all known persons are younger than x On the other hand this universal quantification only considers the known individ uals Not that this query would be impossible to formulate without the NAF negation in nRQL since DLs and RacerPro use the OWA in general We already mentioned that the atom has known successor R is in syntactic sugar for project to x x y R and thus neg has known successor R is equiva lent to neg project to x x y R Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 147 6 1 5 Defined Queries nRQL offers a simple macro mechanism for the specification of defined queries A defined query is just a named query its name can then subsequently be used in other queries so its definition is inserted at the place it is used macro expansion full reset gt okay full reset racer read file OKAY nrql user guide examples family racer Vv N defquery mother of x y and x woman x y has child gt mother of N describe definition mother of defque
130. ar techniques if you are concerned about data privacy and security JRacer JRacer is the client library to access the services of a RacerPro server from Java client programs The package jracer is provided with source code in your RacerPro distribution directory See http www racer systems com products download for the latest version An example client is provided with source code as well The main idea of the socket interface is to open a socket stream submit declarations and queries using strings and to parse the answer strings provided by RacerPro The following code fragment explains how to send message to a RacerPro server running on localhost 127 0 0 1 under port 8088 Racer Systems GmbH amp Co KG http www racer systems com 2 2 THE RACERPRO SERVER 21 import jracer import java io public class Test public static void main String argv String ip 127 0 0 1 int port 8088 String filename Users rm Desktop Examples family owl RacerServer raceri new RacerServer ip port try FileInputStream fstream new FileInputStream filename int n 0 while fstream available 0 int c fstream read n 1 fstream close racerl openConnection System out println raceri send owl read file filename System out println raceri send all atomic concepts raceri closeConnection catch Exception e e printStackTrace We assume that you com
131. are constructed Each query has to be parsed compiled and is then maintained as a query object Since the structure of the subqueries does not change during query execution it would be better to construct these subqueries in advance This can be achieved using so called promises During the time of query preparation a promise for a variable declares that this variable is treated as an individual Moreover it is promised that at the time when the query is executed that variable is indeed bound to an individual Thus a more efficient version looks as follows see file nrql user guide examples lambda ex2 query2 racer with future bindings car delete all queries prepare abox query part and car car car part has part dont show lambdas t id parts of car prepare abox query lambda weight weight told value if exists weight part and car car car part has part part a weight dont show lambdas t id weights of parts of car This has defined two queries named parts of car and weights of parts of car The query optimizer has treated the car variable as an individual due to with future bindings Thus we have promised nRQL that we will only execute these queries if we supply a binding for agg We can establish such a binding during query exe cution using with nrql settings as follows retrievel car car lambda car with nrql settings dont show lambdas t bindings car car
132. arity as the requested tuples Please consider the following example session to get familiar with the QBox facility full reset gt okay full reset racer read file home mi wessel nrql user guide examples family no signature racer gt 0KAY enable query repository gt okay query repository enabled retrieve x x top gt x alice x betty x charles x doris x eve retrieve a a human gt a eve a doris a charles a betty a alice retrieve b b woman gt b alice b betty b doris b eve retrieve y and x woman x y has child y human gt y doris y eve y betty y charles show qbox for abox Racer Systems GmbH amp Co KG http www racer systems com 228 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 333 QBOX FOR racer dummy substrate FOR ABOX smith family 333 O master top query 55 l 4 query 9 333 ___0 master bottom query 55 1 query 1 335 ___2 query 2 SUBQUERY 3 0F query 9 aoe ___3 query 5 SUBQUERY 1 0F query 9 335 ___0 master bottom query gt see output on stdout show qbox for abox smith family t 333 QBOX FOR racer dummy substrate FOR ABOX smith family 333 master top query sais and x ano1 woman x anol y has child y human 333 ___master bottom query i x top rei ___ a human
133. ase note that for macros in contrast to functions the arguments should not be quoted Racer Systems GmbH amp Co KG http www racer systems com 56 CHAPTER 3 RACERPRO KNOWLEDGE BASES C CN top bottom not C and Ci Ch or Ci Ch some R C all R C at least n R at most n R exactly n R at least n R C at most n R C exactly n R C a AN an AN no AN CDC R RN inv RN Figure 3 1 RacerPro concept and role terms Furthermore if an argument tbor or abox is specified in this documentation a name a symbol can be used as well Functions and macros are only distinguished in the Lisp version Macros do not evaluate their arguments If you use the RacerPro server you can use functions just like macros Arguments are never evaluated 3 2 Concept Language The content of RacerPro T boxes includes the conceptual modeling of concepts and roles as well The modeling is based on the signature which consists of two disjoint sets the set of concept names C also called the atomic concepts and the set R containing the role names Starting from the set C complex concept terms can be build using several operators An overview over all concept and role building operators is given in Figure 3 1 The signature does not have to be specified explicitly in RacerPro knowledge bases the system can compute it from the all the used names in the knowledge base but specifying a sign
134. at no individuals appear in NAF negated atoms Use only variables in NAF negated atoms Additional same as and NAF negated same as atoms may be added to further con strain the bindings of variables to certain individuals or to rule out undesired bindings for example retrieve x and x grandmother same as x alice gt x alice describe query last gt query 24 accurate processed Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 133 retrieve x and same as x alice x grandmother abox smith family N retrieve x and meg x grandmother same as x alice gt NIL N describe query last gt query 25 accurate processed retrieve x and same as x alice not x grandmother abox smith family Also the bind individual atom can be used which is syntactic sugar retrieve alice and neg alice grandmother bind individual alice gt NIL N describe query last gt query 10 accurate processed retrieve alice or and same as alice alice not same as alice alice and same as alice alice not alice grandmother top alice abox smith family Please note that NAF negation can also be tricky in the presence of individual synonyms since the standard individual transformation uses same as atoms and same as works on a syntactic level unless you tur
135. ata predicates The recognized predicates are summarized on Page 193 see rule lt data query predicate gt Here is query which uses a data predicate retrieve x x age predicate lt 40 gt NIL retrieve x x age predicate lt 60 gt x betty Whereas the item age is interpreted as before the conjunct list entry predicate lt 60 matches those nodes whose labels contain a numeric literal that satisfies the predicate Predicates can be used in node atoms as well as in edge atoms Please note that the predicate expressions plays the same role as a literal in the and or list structure retrieve x x age predicate 50 predicate 80 gt x betty x alice retrieve Pax x age predicate 50 predicate 80 gt NIL So these queries are somehow equivalent to retrieve x x age 50 80 gt x betty x alice retrieve x x age 50 80 gt NIL Another example demonstrates that predicates are available which are not offered in the concrete domain of RacerPro for example predicates which can search for substrings 18 retrieve x x predicate search Bet 18 gt x betty Moreover it is even possible to find pairs of nodes x and y which satisfy a certain predicate even if there is no explicit edge between these nodes Racer Systems GmbH amp
136. ata substrate label gt lt data substrate label gt gt lt conjunction of data items gt lt conjunction of data items gt gt lt data literal gt lt disjunction of data items gt lt disjunction of data items gt gt lt data literal gt lt data literal gt lt data literal gt gt lt symbol data literal gt lt character data literal gt Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 199 lt string data literal gt lt numeric data literal gt lt symbol data literal gt gt a Common LISP symbol e g symbol lt character data literal gt gt a Common LISP character e g a lt string data literal gt gt a Common LISP string e g string lt numeric data literal gt gt a Common LISP number e g 123 3 2 123 23 Syntax of nRQL data substrate queries lt data node query expression gt gt lt conjunction of data query items gt lt data edge query expression gt gt lt conjunction of data query items gt lt data edge query satisfies expression gt gt satisfies lt 2 ary data query predicate gt lt disjunction of data query predicates gt pA lt data query item gt gt lt data literal gt lt data query predicate gt lt conjunction of data query items gt gt lt data query item gt lt disjunction of data query items gt lt disjunction of data query items gt gt lt data quer
137. atom this object is thus either an ABox individual or a ABox variable For example x and x are variables and betty is an individual Variables are bound to those ABox individuals that satisfy the query expression nRQL offers ordinary variables as well as so called injective variables e Variables are prefixed with e g x y e Injective variables are prefixed with e g x y An injective variable can only be bound to an ABox individual which is not already bound to another injective variable the mapping from variables to ABox individuals is thus injective For exam ple if x is bound to betty then y cannot be bound to betty as well Note that this can also achieved by means of neg same as x y if ordinary variables are used There are only four types of atoms available e Unary atoms concept query atoms e Binary atoms role query atoms constraint query atoms Racer Systems GmbH amp Co KG http www racer systems com 94 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL SAME AS query atoms In principle there is a fifth kind of atom the query reference atom which is n ary and used to refer to a defined query However this atom should be understood on a syntactic macro level and is thus not discussed here So this atom is in fact syntactic sugar as some other additional auxiliary atoms as well e g the bind individual atoms We will now d
138. ature may help avoiding errors caused by typos Racer Systems GmbH amp Co KG http www racer systems com 3 2 CONCEPT LANGUAGE 57 CDC min AN integer max AN integer equal AN integer equal AN AN divisible AN cardinal not divisible AN cardinal gt aexpr aexpr gt aexpr aexpr lt aexpr aexpr lt aexpr aexpr lt gt aexpr aexpr aexpr aexpr string AN string string lt gt AN string string AN AN string lt gt AN AN string gt letter aexpr AN AN must be of type real real aerpri aexpr1 aexpr1 Figure 3 2 RacerPro concrete domain concepts and attribute expressions Boolean terms build concepts by using the boolean operators DL notation RacerPro syntax Negation C not C Conjunction C1 N Chn and Ci Ch Disjunction Cy U UC or Cy Chn Qualified restrictions state that role fillers have to be of a certain concept Value restric tions assure that the type of all role fillers is of the specified concept while exist restrictions require that there be a filler of that role which is an instance of the specified concept DL notation RacerPro syntax 4 R C some R C Y R C all R C Exists restriction Value restriction Number restrictions can specify a lower bound an upper bound or an exact number for the amount of role fillers each instance of this concept has for a certain role
139. benchmark universities departments 8 and x http www lehigh edu X7Ezhp2 2004 0401 univ bench owl Student y Ihttp www lehigh edu 7Ezhp2 2004 0401 univ bench owl Department x y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl memberOf y Ihttp www University0 edul lnttp www lehigh edu X7Ezhp2 2004 0401 univ bench owl subOrganization0Of x a http www lehigh edu X7Ezhp2 2004 0401 univ bench owl emailAddress x y datatype fillers Ihttp www lehigh edu X7Ezhp2 2004 0401 univ bench owl emailAddress x 333 Query 9 benchmark universities departments 9 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Student y Ihttp www lehigh edu X7Ezhp2 2004 0401 univ bench owl Faculty z http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Course x y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl advisor x z http www lehigh edu 7Ezhp2 2004 0401 univ bench owl takesCourse y z http www lehigh edu X7Ezhp2 2004 0401 univ bench owl teacher0Of x y z 333 Query 10 benchmark universities departments 10 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Student x http www Department0O University0 edu GraduateCourse0 http www lehigh edu 7Ezhp2 2004 0401 univ bench owl takesCourse x Racer Systems GmbH amp Co KG http www racer systems com 256 APPENDIX D L
140. berp oddp plusp position princ rationalize remove rest reverse round search second set difference set equal set subset seventh signum sin sinh sixth sqrt string capitalize string downcase string equal string greaterp string lessp string upcase string lt string lt string string gt string gt stringp symbol name symbolp tan tanh tenth terpri third tree equal type of union with open output file write zerop 6 1 3 2 MiniLisp Head Projection Operators We have already discussed the told value and attribute retrieval head projection operators nRQL offers some more head projection operators which are im plemented in MiniLisp The following head projections operators are imple mented in MiniLisp and can be called MiniLisp macro head projection operators all types also types instantiators all instantiators all types flat also all instantiators flat types flat instantiators flat and direct types also most specific types most specific instantiators direct instantiators direct types flat also most specific types flat most specific instantiators flat direct instantiators flat and describe as well as individual synonyms Here are some examples illustrating these macro operators full reset gt okay full reset N instance i c gt 0KAY implies c d gt 0KAY retrieve x describe x x top x i lambda x describe ind x current abox x i assertions Ci c r
141. bers characters and even strings We easily verify that betty is an instance in the substrate as well as in the ABox all concept assertions gt betty and woman age 50 get data node description betty gt node name betty node label Betty age 50 Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 169 node successors NIL node predecessors NIL describe all nodes gt node name betty node label Betty age 50 node successors NIL node predecessors NIL describe substrate gt type data substrate abox default no of nodes 1 no of edges 0 Note that betty in the ABox and betty in the substrate are associated since the same name is used However it is possible to create substrate individuals without associated ABox individual simply don t supply a concept a third argument to data node In general RacerPro s reasoning is completely unaffected by the presence of an associated data substrate layer Thus you might ask what is it good for at all The answer is You can use nRQL to query the hybrid representation nRQL is a hybrid query language To demonstrate this we create one more node as well as a data substrate edge data node alice Alice age 80 and woman age 80 some has child human gt alice data edge alice betty has child Alice this mother of Betty has
142. bject for which an individual stands a member of the set of objects described by a specified concept The individual is then called an instance of the concept e Instance retrieval w r t an A box and a T box Find all individuals from an A box such that the objects they stand for can be proven to be a member of a set of objects described by a certain query concept e Retrieval of tuples of individuals instances that satisfy certain conditions w r t an A box and a T box e Computation of the direct types of an individual w r t an A box and a T box Find the most specific concept names from a T box of which a given individual is an instance e Computation of the fillers of a role with reference to an individual w r t an A box and a T box e Check if certain concrete domain constraints are entailed by an A box and a T box Racer Systems GmbH amp Co KG http www racer systems com 4 CHAPTER 1 INTRODUCTION RacerPro provides another semantically well defined query language nRQL new Racer Query Language which also supports negation as failure numeric constraints w r t at tribute values of different individuals substring properties between string attributes etc In order to support special OWL features such as annotation and datatype properties special OWL querying facilities have been incorporated into nRQL The query language OWL QL 4 is the W3C recommendation for querying OWL documents nRQL has been used as a ba
143. ble year 4 or not divisible year 100 divisible year 400 equivalent February and Month lt days in month 29 or not month inleapyear days in month 29 or month inleapyear days in month 28 Next we assume some instances of February are declared instance feb 2003 February constrained feb 2003 year 1 year constrained feb 2003 days in feb 2003 days in month constraints year 1 2003 instance feb 2000 February constrained feb 2000 year 2 year constrained feb 2000 days in feb 2000 days in month constraints year 2 2000 Note that the number of days for both months is not given explicitly Nevertheless asking concept instances month inleapyear yields feb 2000 whereas asking for concept instances not month inleapyear returns feb 2003 In addition one could check the number of days constraint entailed lt gt days in feb 2003 29 constraint entailed days in feb 2000 29 In both cases the answer is true Racer Systems GmbH amp Co KG http www racer systems com 3 6 CONCRETE DOMAIN ATTRIBUTES 65 3 6 Concrete Domain Attributes Attributes are considered as typed since they can either have fillers of type cardinal integer real complex or string The same attribute cannot be used in the same T box such that both types are applicable e g min has age 18 and gt has age 18 are not allowed If the type of an attribute is not explici
144. bound to ABox individuals However an ABox may also contain so called concrete domain objects CD objects as well as told values data values literals from the concrete domain We already showed that a concept query atom such as x gt age 18 can be used to specify retrieval conditions on concrete domain attributes here age However no variable can be bound to the age attribute value even if there is a told value given in the ABox Thus head projection operators are provided in order to retrieve concrete domain values and or concrete domain objects full reset okay full reset v racer read file nrql user guide examples family racer gt 0KAY N retrieve x x gt age 18 x alice x doris x eve x betty x charles v N retrieve x told value if exists age x x gt age 18 x alice existing told values age x 80 x doris existing told values age x 24 x eve existing told values age x 18 x betty existing told values age x 50 x charles existing told values age x 55 Vv Please note that the concrete age cannot always be determined In this example the told values have been gather from the concept assertions in family racer of the form instance instance alice age 80 etc nRQL also considers constrained and constraints assertion in the ABox to identify told values for example
145. bstrate nRQL is a hybrid query language If the data substrate is enabled then the so called data substrate query atoms can be used to query the data substrate associated with an ABox The data substrate of an ABox is a representation layered on top of an ABox A full set of retrieval predicates can be evaluated on that representation substring searching etc Racer Systems GmbH amp Co KG http www racer systems com 90 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL So called complex TBox queries are available These can be used to search for certain patterns of sub superclass relationships in a taxonomy Optionally nRQL is a hybrid query language which can query a hybrid representation consisting of an ABox associated layered with a so called substrate To query such a hybrid ABox substrate representation a hybrid nRQL query may be used which contains query atoms which are evaluated on the ABox as well as query atoms which are evaluated on the substrate The language offers extensibility and flexibility by means of a simple expression lan guage called MiniLisp With MiniLisp a user can write a simple termination safe program which is executed on the RacerPro server MiniLisp programs give the flex ibility to specify the format of the query answers write query answers to a file start subqueries etc Moreover and most importantly MiniLisp enables the formulation of aggregation operators count sum aver
146. called thrashing a term borrowed from constraint programming is thus minimized However candidate generation for variable bindings will also be more expensive since the concepts used in concept query atoms will be more complex Thus it depends on the KB and the query if the realized query will perform better than the non realized one Racer Systems GmbH amp Co KG http www racer systems com 234 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 2 10 The nRQL Persistency Facility nRQL maintains a so called substrate data structure for each RacerPro ABox which was queried This ABox associated substrate data structure contains index structures and other auxiliaries needed for query answering Computation of these index structures is triggered on demand if needed The first time a query is posed to a certain ABox which has not seen a nRQL query before nRQL creates a corresponding substrate for the ABox as well as all its index structures Thus answering the first query to an ABox takes considerably longer than subsequent queries to the same ABox In order to avoid recomputation of these index structures it is possible to dump store the complete substrate data structure into a file Moreover also the data substrate or RCC substrate may contain nodes which have been created by hand and which must be saved somehow The API functions store substrate for abox and store substrate for current abox are provided for this purpose
147. ce to RacerPro 2 4 1 Preferences In order to specify preferences simply press the Profiles button Mac choose Prefer ences item from the menu specify the RacerPro host and port to connect to and then Connect to that RacerPro server See also Figure 2 9 Note that you can manage different server settings with profiles The file porter pot in your home directory is used to store your Racer Systems GmbH amp Co KG http www racer systems com 2 4 RACERPORTER 27 profiles The default profile will be automatically used on startup You will be connected automatically if auto connect is specified in the default profile If you are successfully connected you can use the shown tabs to access the functionality of RacerPro The six text fields at the top of the application are called the state display of RacerPorter They indicate the current state of RacerPorter NOT of RacerPro Selecting for example a TBox from the TBoxes list panel updates the TBox field in the state display The same applies to the other fields of the state display The state is used to indicate the current object which is to be inspected or queried with RacerPorter for example the Individuals tab will always show the individuals of the ABox in the state display the Taxonomy tab will always show the taxonomy of TBox in the state display etc Note that selecting an ABox TBox from the ABoxes TBoxes pane does NOT change the current abo
148. cepts top and bottom Racer Systems GmbH amp Co KG http www racer systems com 3 3 CONCEPT AXIOMS AND T BOXES 59 Concrete domain concepts state concrete predicate restrictions for attribute fillers see Figure 3 2 RacerPro currently supports three unary predicates for integer attributes min max equal six nary predicates for real attributes gt gt lt lt lt gt a unary existen tial predicate with two syntactical variants a or an and a special predicate restriction disallowing a concrete domain filler no The restrictions for attributes of type real have to be in the form of linear inequations with order relations where the attribute names play the role of variables If an expression is built with the rule for aexpr4 see Figure 3 2 a so called nonlinear constraint is specified In this case only equations and inequations lt gt but no order constraints gt gt lt lt are allowed and the attributes must be of type complex If an expression is built with the rule for aexpr5 see Figure 3 2 a so called cardinal linear constraint is specified i e attributes are constrained to be a natural num ber including zero RacerPro also supports a concrete domain for representing equations about strings with predicates string and string lt gt The use of concepts with concrete domain expressions is illustrated with examples in Section 3 5 For the declaration of types for attributes se
149. child gt alice betty Note that data edge accepts similar to data node an optional argument in case of data node the optional argument is a RacerPro concept which adds a concept assertion to the associated ABox and in case of data edge a role assertion is created Note that we have attached the label has child Alice this mother of Betty to the edge and used the role has child for the role assertion Thus describe all edges gt from node alice to node betty Racer Systems GmbH amp Co KG http www racer systems com 170 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL edge label has child Alice is mother of Betty get data edge description alice betty gt from node alice to node betty edge label has child Alice this mother of Betty gt all role assertions gt alice betty has child all concept assertions gt alice and woman age 80 some has child human betty and woman age 50 some has child human node label alice gt Alice age 80 edge label alice betty gt has child Alice is mother of Betty Let us add some trivial background knowledge to make the following queries more interest ing implies woman human gt OKAY define concept mother and woman some has child top gt OKAY Please note that nodes and edges can also be deleted see delete data node delete data edge Please
150. chniques in order to speedup answering subscription based instance retrieval queries such that reasonably large set of documents can be handled Furthermore it will be possible to define how applications are to be informed about matches to previous subscriptions i e event handlers can be introduced 5 3 2 Using JRacer for Publish and Subscribe The following code fragment demonstrates how to interact with a RacerPro server from a Java application The aim of the example is to demonstrate the relative ease of use that such an API provides In our scenario we assume that the agent instructs the Rac erPro system to direct the channel to computer rm sts tu harburg de at port 8080 Before the subscription is sent to a RacerPro server the agent should make sure that at rm sts tu harburg de the assumed agent base station a so called listener process is started at port 8080 This can be easily accomplished by starting the following program on rm sts tu harburg de Racer Systems GmbH amp Co KG http www racer systems com 5 3 THE PUBLISH SUBSCRIBE MECHANISM 85 public class Listener public static void main String argv try ServerSocket server new ServerSocket 8080 while true Socket client server accept BufferedReader in new BufferedReader new InputStreamReader client getInputStream String result in readLine in close catch IOException e If a message comes in ov
151. clared as R REIR element of F Va aa F Figure 3 4 Conflicting declared and inherited role properties The combination of a feature having a transitive superrole is not allowed and features cannot be transitive Note that transitive roles and roles with transitive subroles may not be used in number restrictions RacerPro does not support role terms as specified in the KRSS How ever a role being the conjunction of other roles can as well be expressed by using the role hierarchy cf 5 The KRSS like declaration of the role define primitive role RN and RN RNo2 can be approximated in RacerPro by define primitive role RN parents RN RN2 KRSS DL notation define primitive role RN domain C A RN T CC define primitive role RN range D TC Vv RN D RacerPro Syntax DL notation define primitive role RN domain C A RN T CC define primitive role RN range D TC Vv RN D Figure 3 5 Domain and range restrictions expressed via GCIs RacerPro offers the declaration of domain and range restrictions for roles These restrictions for primitive roles can be either expressed with GCIs see the examples in Figure 3 5 cf 5 or declared via the keywords domain and range 3 5 Concrete Domains RacerPro supports reasoning over natural numbers N integers Z reals R complex numbers C and strings For different sets different kinds of predicates ar
152. concepts C and adds implied concept assertions of the form instance i D if D is implied by C w r t the TBox In order to find appro priate D to add nRQL classifies the TBox computes the taxonomy and uses the atomic concept synonyms and atomic concept ancestors of C for D For example if the ABox contains the concept assertion instance betty mother then also instance betty woman will be added since woman is a member of atomic concept ancestors mother Moreover conjunctions are broken up like in Mode 0 and the same process is applied recursively until no more assertions can be added The resulting upward saturated ABox is then queried like in the previous setting Note that in contrast to the previous setting retrieve x x woman now correctly returns x betty However this mode will still fail for the query retrieve x x and woman human This is what the next mode achieves In order to enable this mode use set nrql mode 1 Told information querying plus taxonomic information for all concepts Mode 2 This mode is like Mode 1 but now the atomic concept synonyms and ancestors will also be computed for arbitrary and not only atomic concepts C in concept membership assertions instance i C Thus if the ABox contains instance betty or woman mother then RacerPro will compute the equivalent concepts and concept ancestors of the compound non atomic concept or woman mother This will r
153. ct gt has known successor has child has parent has descendant has ancestor has synonym Racer Systems GmbH amp Co KG http www racer systems com 198 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL nRQL data substrate query language lt data query atom gt gt not neg lt data query atom gt lt data query object gt lt data node query expression gt lt data query object gt lt data query object gt lt data edge query expression gt lt data query object gt nil lt data edge query expression gt nil lt data query object gt lt data edge query expression gt top lt data query object gt bottom lt data query object gt lt data query object gt lt data query object gt lt data edge query satisfies expression gt bind individual lt data query individual gt wo same as won equal lt data query object gt lt data query object gt lt data query object gt has known successor lt data edge query expression gt Substitute insert lt query name gt lt query object gt lt data query object gt nil lt query object gt lt data query object gt nil lt query name gt Syntax of labels of data nodes and edges used in data node lt symbol gt lt data substrate label gt data edge lt symbol gt lt d
154. ctively use RacerPro load knowledge bases visualize on tologies T Boxes and instances A Boxes send queries and native commands to RacerPro Start RacerPro and start RacerPorter Then in RacerPorter press the button Connect Racer Systems GmbH amp Co KG http www racer systems com 26 CHAPTER 2 USING RACERPRO Enter your commands in RacerPro native syntax in the Command pane The replies of RacerPro are put into the RacerLog pane Just hit Return in case you need more than one line for your command The shell will not send the input to Racer until the last parenthesis has been closed If you press enter on an incomplete expression RacerPorter just gives you a fresh line starting with appropriate indentation 9600 RacerPorter f Shell Cons TBoxes ABoxes T Choose Profile _Localhost HH Indivi Asser Queries Rules TBox Profile Name Localhost Concept Individual RacerPro TCP IP Address localhost RacerPro Port Number 8088 Command _ Auto Connect v Make Default Profile Apply New Delete Save OK Cancel Connect Profiles Cle I ll Reset New Open j RacerPro Log Do This is RacerPorter Version 1 9 0 With RacerPorter you can interactivly use RacerPro Load knowledge bases visualize ontologies T Boxes and instances A Boxes send queries and native commands to the server Figure 2 9 The RacerPorter inferfa
155. d at least 1 has gender a has age implies human temperature fahrenheit equivalent equivalent equivalent equivalent equivalent equivalent 1 8 temperature celsius 32 young human and human max has age 20 teenager and young human min has age 10 adult human and human min has age 21 old human and human min has age 60 woman and human all has gender female man and human all has gender male implies child human equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent equivalent young child and child max has age 9 human with fever and human gt temperature celsius 38 5 seriously ill human and human gt temperature celsius 42 0 human with high fever and human gt temperature fahrenheit 107 5 parent at least 1 has child young parent and parent max has age 21 normal parent and parent min has age 22 max has age 40 old parent and parent min has age 41 mother and woman parent father and man parent teenage mother and mother max has age 20 mother having only female children and mother all has child all has gender not male mother having only daughters and woman at least 1 has child all has child woman mother with children and mother at least 2 has child
156. d of doris x betty y doris x betty y eve x alice y betty x alice y charles 6 1 1 3 Constraint Query Atoms Constraint query atoms are binary atoms and address the concrete domain part of an ABox Like role query atoms they are used to retrieve pairs of individuals which are in a certain relationship to one another However this relationship is not a role but specified with a possibly complex binary concrete domain predicate like For example we can retrieve those pairs of individuals that have the same age age is a so called concrete domain attribute of type cardinal The family racer KB already contains such an age attribute as well as assertions specifying the individual ages of the family members instance alice age 80 instance betty age 50 instance charles age 55 instance eve age 18 instance doris age 24 Using a concept query atom on can easily identify the adult family members retrieve x x gt age 18 gt x alice x doris x eve x betty x charles However in order to find out who is older than whom a binary retrieval predicate must be used retrieve x y x y constraint age age lt gt x doris y alice x doris y betty x doris y charles x eve y alice x eve y doris x eve y betty x eve y charles x betty y alice x betty
157. d values age x no cd objects x betty told values age x no cd objects x charles told values age x no cd objects Please note that told value if exists is more general 6 1 2 2 Retrieving Told Fillers of OWL Datatype Properties OWL KBs may contain datatype fillers of datatype properties In OWL these fillers are val ues literals of XSD XML Schema datatypes In RacerPro these fillers are represented as told values in the concrete domain part of the ABox With the help of the datatype fillers head projection operator it is possible to retrieve these data fillers Let us consider owl ex3 owl again which contains two datatype properties name and age full reset gt okay full reset owl read file nrgl user guide examples owl ex3 owl gt nrgql user guide examples owl ex3 owl retrieve x datatype fillers name x datatype fillers age x x and person gt age 30 gt x http www owl ontologies com unnamed owl alice datatype fillers http www owl ontologies com unnamed owl name x Alice datatype fillers http www owl ontologies com unnamed owl age x 80 Racer Systems GmbH amp Co KG http www racer systems com 114 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL x http www owl ontologies com unnamed owl betty datatype fillers http www owl ontologies com unnamed owl name x Betty datatype
158. different A boxes are to be tested they should probably be located separately from the associated T boxes in order to save processing time During the development phase of a T box it might be advantageous to call inference services directly For instance during the development phase of a T box it might be useful to check which atomic concepts in the T box are inconsistent by calling check tbox coherence This service is usually much faster than calling classify tbox However if an applica tion problem can be solved for example by checking whether a certain A box is con sistent or not see the function abox consistent p it is not necessary to call either check tbox coherence or classify tbox For all queries RacerPro ensures that the knowledge bases are in the appropriate states This behavior usually guarantees minimum runtimes for answering queries These roles are treated as roles that are neither a feature nor transitive and do not have any superroles New items are added to the T box Note that this might lead to surprising query results e g the set of subconcepts for T contains concepts not mentioned in the T box in any concept axiom Therefore we recommend to use a signature declaration see below SThese concepts are assumed to be atomic concepts Racer Systems GmbH amp Co KG http www racer systems com 3 9 RETRACTION AND INCREMENTAL ADDITIONS 67 3 9 Retraction and Incremental Additions RacerPro offer c
159. e ABox only contains the following role assertions all role assertions gt eve doris has sister doris eve has sister charles betty has sibling betty eve has child betty doris has child alice charles has child alice betty has child Thus the labeling information has sibling has been added according to the role hierarchy since doris eve has sister implies doris eve has sibling 6 1 8 2 2 Markers Used in the Mirror Data Substrate Let us summarize the markers which are used For the nodes e abox object e abox individual e abox concrete domain object e abox concrete domain value For the edges e abox object e abox relationship e abox role relationship e abox attribute relationship e abox told value relationship In addition some additional markers are added in case the ABox has been created from an OWL KB e owl object e 0wl relationship Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 181 e owl individual e owl class e owl literal e owl object property relationship e owl datatype property relationship e owl object annotation property relationship e owl datatype annotation property relationship 6 1 8 2 3 Using the Mirror Data Substrate on OWL Documents Especially in the case of OWL KBs the additional querying functionality provided by the mirror substrate is valuable OWL KBs te
160. e current binding of x So within the lambda car is bound to the bindings value of x First the total weight is computed for this purpose a subquery is constructed If x cari then the query retrievet and cari car cari part has part part a weight is constructed and posed asking for the parts of car1 The head of the subquery consists of yet another lambda which simply applies the told value if exists head projection operator to retrieve the told values of the weight attribute of part The subquery result is returned as a nested list the list if then flattened and its items are summed using reduce The result is bound to the local variable Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 123 w Similarly the number of parts is computed by posing yet another subquery Finally the result of the lambda expression is constructed and returned The constructed and returned value will become the result tuple So if car is car1 and no of parts is 4 and w is 630 0 then the template car car no of parts parts total weight w con structs the result tuple car cari no of parts 4 total weight 630 0 6 1 3 3 2 Efficient Aggregation Operators using Promises Although the previous query demonstrated the power and utility of MiniLisp the aggregation is not computed efficiently The reason is that for each binding of car two new subqueries
161. e query atoms are binary atoms Role query atoms are used to retrieve role fillers from an ABox or OWL RDF S individuals which are realated via a certain OWL object property For example we can retrieve all individuals in the ABox of family racer using the following query whose body consists of a single role query atom x y has child full reset okay full reset v N racer read file nrql user guide examples family racer OKAY v Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 99 retrieve x y x y has child gt x betty y doris x betty y eve x alice y betty x alice y charles The expression mother child has child is a role query atom Again individuals can be used at any position where a variable is accepted to retrieve the children of betty retrieve child of betty betty child of betty has child gt child of betty eve child of betty doris Please note that concrete domain attributes cannot be used in role query atoms Fillers of concrete domain attributes can be retrieved by means of the so called head projection operators see below family racer contains the age attribute of type cardinal retrieve x y x y age gt ERROR Parser Error Unrecognized expression x y age However features functional roles are a special kind of roles and thus can be used for examp
162. e see Section 3 6 DL notation RacerPro syntax Concrete filler exists restriction JA Tp a A or an A No concrete filler restriction VA Lp no A Integer predicate exists restriction JA min min A 2 with z E Z JA max max A z JA equal A z Real predicate exists restriction dA1 An P P aexpr aexpr with P gt gt lt lt An all restriction of the form VA1 An P is currently not directly supported However it can be expressed as a disjunction VA1 Lp U U YAn Lp U JA1 An P 3 3 Concept Axioms and T boxes RacerPro supports several kinds of concept axioms General concept inclusions GCIs state the subsumption relation between two concept terms DL notation Ci E Cy RacerPro syntax implies Ci C2 Concept equations state the equivalence between two concept terms DL notation Cy Co RacerPro syntax equivalent Ci C2 Concept disjointness axioms state pairwise disjointness between several concepts Dis joint concepts do not have instances in common DL notation Cy E 7 C2U C3 U Cn Cz E 7A C3U Cp Racer Systems GmbH amp Co KG http www racer systems com 60 CHAPTER 3 RACERPRO KNOWLEDGE BASES Cai ACh RacerPro syntax disjoint Ci Cn Actually a concept equation Cj C gt can be expressed by the two GCIs Ci E C2 and Cy E Ci The disjointness of the concepts Ci Ch can also be
163. e supported Racer Systems GmbH amp Co KG http www racer systems com 62 CHAPTER 3 RACERPRO KNOWLEDGE BASES N linear inequations with order constraints and integer coefficients Z interval constraints R linear inequations with order constraints and rational coefficients Strings equality and inequality For the users convenience rational coefficients can be specified in floating point notation They are automatically transformed into their rational equivalents e g 0 75 is transformed into 3 4 In the following we will use the names on the left hand side of the table to refer to the corresponding concrete domains Names for values from concrete domains are called objects The set of all objects is referred to as O Individuals can be associated with objects via so called attributes names or attributes for short Note that the set A of all attributes must be disjoint to the set of roles and the set of features Attributes can be declared in the signature of a T box see below The following example is an extension of the family T box introduced above In the example the concrete domains Z and R are used signature atomic concepts teenager roles attributes integer age equivalent teenager and human min age 16 equivalent old teenager and human min age 18 Asking for the children of teenager reveals that old teenager is a teenager A further extensions demonstrates the usage o
164. e to the T box the minimum number of fillers is 3 But should we add at most 1 r or at most 1 r and a b c The first one might be too strong because of i being an instance of some r d What about d Would it be a good idea to also add at most 1 r d If yes then the question arises how to determine the qualifier concepts used in qualified number restrictions 4 4 Closed World Assumption Due to the non monotonic nature of the closed world assumption and the ambiguities about what closing should actually mean in description logic inference systems usually there is no support for the closed world assumption However with the publish and subscribe interface of RacerPro users can achieve a local closed world LCW assumption see Section 5 3 The nRQL query language allows you to query A boxes using negation as failure semantics see Chapter 6 4 5 Unique Name Assumption In addition to the Open World Assumption it is possible to instruct RacerPro to employ the Unique Name Assumption UNA This means that all individuals used in an A box are assumed to be mapped to different elements of the universe i e two individuals cannot refer to the same domain element Hence adding instance alice at most 1 has child does not identify betty and charles but makes the A box inconsistent Due to our experi ence with users we would like to emphasize that most users take UNA for granted but are astonished to learn that OWA is assumed rather
165. e when the first query is answered The function abox consistent is not to be called if set nrql mode 1 is called see above The following sequence of statements provides for fastest execution times Reasoning is only complete for ALE with simple GCIs and A boxes for which there exists a filler for every exists restriction offline phase full reset set nrql mode 1 enable optimized query processing set unique name assumption t time load data time prepare abox time prepare racer engine online phase retrieve Nevertheless it might be useful to use this mode for information retrieval problems see also Chapter 6 5 2 The RacerPro Persistency Services If you load some knowledge bases into RacerPro and ask some queries RacerPro builds internal data structure that enables the system to provide for faster response times However generating these internal data structures takes some time So if the RacerPro server is shut down all this work is usually lost and data structures have to be rebuilt when the server is restarted again In order to save time at server startup RacerPro provides a facility to dump the server state into a file and restore the state from the file at restart time The corresponding functions form the persistency services of a RacerPro server The persistency services can also be used to prepare a knowledge base at a specific server and use it repeatedly at multiple
166. e wondering why a certain query never returns any tuples and you think it should you should consider using the consistency checking service for queries An inconsistent query must necessarily produce an empty answer on all ABoxes thus CPU cycles can be saved if such queries are recognized in advance before executing them Even though the reasoning mechanism offered by nRQL for queries are incomplete and yet experimental they are still useful The consistency checking service is sound and complete for queries which do contain at most one occurrence of neg The idea of the algorithm is simple The query is first transformed into DNF Then for each disjunct a canonical ABox is constructed witnessing the satisfiability of the query Note that each disjunct is a grounded conjunctive query Please note that the ABox is not taken into account e g even if betty is an instance of woman and woman and man are disjoint then the query retrieve betty man is not recognized as inconsistent The reasons is that in principle there could be and ABox in which betty is indeed a man Here is an example session demonstrating the utility of nRQL s reasoning facilities full reset okay full reset v racer read file home mi wessel nrql user guide examples family no signature racer gt OKAY report inconsistent queries and rules gt okay reporting inconsistent queries and rules retrieve x and x woman x man
167. eight lambda weight weight told value if exists weight part parts length retrieve part car part has part format output stream Car with name A weights A kg and has CA parts car w parts car gt NIL Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 125 The output can be found in the file minilisp output txt moreover the result NIL is delivered because the lambda body returns NIL Please note that with open output file only works if RacerPro in running in unsafe mode 6 1 3 3 5 Filtering Result Tuples lambda bodies can also work as filters as follows if the special token reject is returned from the lambda body then the result is rejected i e will not appear in the query answer For example using reject we can easily reject all parts which have no subparts retrieve1 part top lambda part let parts length retrieve part part part has part if zerop parts reject part has parts parts part gt enginet has 1 parts cari has 4 parts 6 1 3 3 6 Combined TBox ABox Queries Finally let us present an example with demonstrates how to combine TBox queries and ABox queries Sometimes one wants to retrieve all and only the direct instances of a concept OWL class An individual is called a direct instance of a concept OWL class if there is no subconcept subcla
168. eiiseoioniesocoseeeeeeeeeioooeoosete B Welcome to RacerPorter the graphical user interface for RacerPro Seeeeseseeseesesseseesessesseseesessessessess Seeseseesesseseesessesss This is RacerPorter Version 1 9 0 With RacerPorter you can interactivly use RacerPro load knowledge bases visualize ontologies T Boxes and instances A Boxes send queries and native commands to the server Figure 2 1 The RacerPorter interface for RacerPro 2We assume that RacerPorter is on the search path of your operating system Racer Systems GmbH amp Co KG http www racer systems com 2 1 SAMPLE SESSION WITH RACERPORTER 13 RacerPorter also allows you to edit knowledge bases or ontologies and data descriptions with the built in editor Just press the button Open or select Open in the File menu Mac OS X Under Mac OS X and Windows you can also double click the file family racer make sure RacerPro is already running If you have the file family racer displayed in the editor see Figure 2 2 you can send the statements to RacerPro by selecting Evaluate Racer Buffer in the Buffer menu make sure the window RacerEditor is the active window 600 RacerEditor Cin knowledge base family smith family Csignature Cimplies Cimplies Cimplies Cimplies Cimplies Cimplies Cimplies Chuman person female male woman man parent mother father grandmother aunt uncle sister brother es has descendan
169. eleases of RacerPro subscriptions can be extended with information about how the retrieval events are to be signaled to the client This will be done with a proxy which is currently under development The example is continued with the following statements and two new subscriptions state instance document 6 computer science document constrained document 6 isbn 6 isbn constraints equal isbn 6 6234567 publish document 6 state subscribe client 2 computer science document subscribe client 3 computer science best seller The last statement returns two additional pairs indicating the retrieval results for the in stance retrieval query subscription of client 2 CLIENT 2 DOCUMENT 5 CLIENT 2 DOCUMENT 6 Next information about another document is declared The new document is published state instance document 7 computer science document constrained document 7 isbn 7 isbn constraints equal isbn 7 7234567 constrained document 7 number of copies sold 7 number of copies sold constraints equal number of copies sold 7 4000 publish document 7 Racer Systems GmbH amp Co KG http www racer systems com 82 CHAPTER 5 KNOWLEDGE BASE MANAGEMENT The result of the last statement is CLIENT 2 DOCUMENT 7 CLIENT 3 DOCUMENT 7 The new document document 7 is in the result set of the query subscribed by client 2 and client 3 Note that document can be considered as structured objects not ju
170. ence gt gt lt query body gt lt generalized abox assertion gt gt lt instance assertion gt lt related assertion gt lt same as assertion gt lt constrained assertion gt lt constraints assertion gt lt forget concept assertion gt lt forget role assertion gt lt forget constrained assertion gt lt forget constraint assertion gt Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 201 lt generalized object gt gt lt query object gt lt new ind operator gt l lt abox head projection operator gt lt lambda application gt lt new ind operator gt gt new ind new individual individual create individual new symbol new atom new concept lt symbol gt lt query object gt lt concept expression2 gt gt a RacerPro concept expression which may contain lt new ind operator gt allowing for the construction of new atomic concepts based on bindings of lt query objects gt lt instance assertion gt gt instance lt generalized object gt lt concept expression2 gt lt related assertion gt gt related lt generalized object gt lt generalized object gt lt role expressioni gt lt same as assertion gt gt same as lt generalized object gt lt generalized object gt lt constrained assertion gt gt constrained lt generalized object gt lt generalized objec
171. ent from Lispworks Ltd Racer Systems GmbH amp Co KG http www racer systems com 1 4 ACKNOWLEDGMENTS 7 http www lispworks com RacerPro is also developed with Macintosh Common Lisp http www digitool com The XML RDF based part of the input interface for RacerPro is implemented using the XML RDF parser Wilbur written by Ora Lassila For more information on Wilbur and the Wilbur source code see http wilbur rdf sourceforge net For most versions of RacerPro the DIG server interface is implemented with Al legroServe For some special versions of RacerPro however the HTTP server for the DIG interface of RacerPro is implemented with CL HTTP which is devel oped and owned by John C Mallery For more information on CL HTTP see http www ai mit edu projects iiip doc cl http home page html Many users have directly contributed to the functionality and stability of the RacerPro system by giving comments providing ideas and test knowledge bases implementing inter faces or sending bug reports There are too many of them to mention the name them all Many thanks for any hint comment and bug report Racer Systems GmbH amp Co KG http www racer systems com CHAPTER 1 INTRODUCTION Racer Systems GmbH amp Co KG http www racer systems com Chapter 2 Using RacerPro In this section we present a first example for the use of RacerPro We use the so called KRSS based interface rather tha
172. er lt rdf RDF xmlns rdf http www w3 0rg 1999 02 22 rdf syntax ns xmlns xsd http www w3 org 2001 XMLSchema xmlns rdfs http www w3 org 2000 01 rdf schema xmlns owl http www w3 org 2002 07 owl xmlns http www owl ontologies com unnamed owl xml base http www owl ontologies com unnamed owl gt lt owl Ontology rdf about gt Note that has gender has been declared as a so called feature a functional role So any person can have at most one gender and male and female are disjoint Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 101 lt owl Class rdf ID person gt lt owl ObjectProperty rdf ID hasChild gt lt rdfs domain rdf resource person gt lt rdfs range rdf resource person gt lt owl O0bjectProperty gt lt person rdf ID alice gt lt hasChild gt lt person rdf ID betty gt lt hasChild gt lt person gt lt rdf RDF gt Thus there are two instances of class person The individual michael is the filler of the hasChild object property of the individual margrit Since object properties are mapped to roles in RacerPro role query atoms can be used full reset gt okay full reset owl read file 2 owl read file nrql user guide examples owl ex2 owl gt nrql user guide examples owl ex2 owl N retrieve x y x y hasChild x http www owl ontologies com unnamed owl alice y ht
173. er However if we refer to this concept using its name we can retrieve all concepts which have a name different from mother tbox retrieve mother neg mother mother gt mother bottom mother bottom mother woman mother woman with children mother top mother top Please recall that the standard transformation replaces variables with representative vari ables 6 1 7 3 How to Retrieve All Concept Names The query x top will retrieve those nodes which satisfy top Unlike in a real ABox there are only two nodes which satisfy top tbox retrieve x x top gt x top x top To retrieve all individuals from the taxonomy ABox use either the special top atom or if you only want to retrieve the representative concepts use taxonomy node as already demonstrated Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 165 tbox re gt x bo x x x x x x tbox re gt x to 6 1 7 4 C trieve x top x ttom bottom woman mother woman with children top top trieve x x taxonomy node p x woman with children x woman x bottom omplex TBox Queries Finally let us consider a more complex query which searches for for diamond shaped super subclass relationships in the taxonomy Thus we are looking for a node x that has descen
174. er the input stream the variable result is bound accordingly Then the message can be processed as suitable to the application We do not discuss details here The subscription to the channel i e the registration of the query can also be easily done using the JRacer interface as indicated with the following code fragment we assume RacerPro runs at node racer racer systems com on port 8088 public class Subscription public static void main String argv RacerServer raceri new RacerServer www racer systems com 8088 String res try raceri openConnection res raceri send subscribe q_1 Book notification method tcp rm sts tu harburg de 8 raceri closeConnection System out println res catch Exception e e printStackTrace The connection to the RacerPro server is represented with a client object of class RacerSocketClient The client object is used to send messages to the associated Rac erPro server using the message send Control flow stops until RacerPro acknowledges the subscription 5 3 3 Realizing Local Closed World Assumptions Feedback from many users of the RacerPro system indicates that for instance instance retrieval queries could profit from possibilities to close a knowledge base in one way or Racer Systems GmbH amp Co KG http www racer systems com 86 CHAPTER 5 KNOWLEDGE BASE MANAGEMENT another Due to the non monotonic nature of the closed
175. es not represent meta level knowledge in the theory because this might result in paradoxes which are reported else where The triples in RDFS files are processed in a special way They are represented as T box axioms If the property is rdf type the object must be rdfs Class or rdfs Property These statements are interpreted as declarations for concept and role names respectively Three types of axioms are supported with the following properties rdfs subClass0f rdfs range and rdfs domain Other triples are ignored OWL files are processed in a similar way The semantics of OWL is described elsewhere see http www w3 org TR owl ref There are a few restrictions in the RacerPro implemen tation The UNA cannot be switched off and number restrictions for attributes datatype properties are not supported Only basic datatypes of XML Schema are supported i e RacerPro cannot read files with datatype declarations right now Usually the default namespace for concept and role name is defined by the pathname of the OWL file If the OWL file contains a specification for the default namespace i e a specification xmlns this URL is taken as a prefix for concept and role names lt xml version 1 0 encoding IS0 8859 1 gt lt rdf RDF xmlns http www mycompany com project xmlns owl http www w3 org 2002 07 owl xmlns rdf http www w3 0rg 1999 02 22 rdf syntax ns xmlns rdfs http www w3 0rg 2000 01 rdf schema xm
176. eset instance alice grandmother gt OKAY instance betty and woman some has child top v OKAY implies grandmother mother v OKAY define concept mother and woman some has child top gt N OKAY set nrql mode 0 gt 0kay mode 0 Vow retrieve x x woman x betty Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 215 set nrql mode 1 okay mode 1 Vv retrieve x x woman gt x alice x betty retrieve x x mother gt x alice set nrql mode 3 gt okay mode 3 retrieve x x mother gt x betty x alice 6 2 5 3 The Two Phase Query Processing Modi The so called two phase query processing modi are special lazy incremental tuple at a time modi The modi 4 5 and 6 are two phase processing modi All modi are complete Mode 4 and 5 are lazy incremental modi whereas mode 6 is like mode 3 a set a t time mode Let us describe the modi 4 and 5 first If nRQL is used in these modi then delivery of tuples will be arranged in two phases e In phase one the so called cheap tuples will be delivered e followed by the expensive tuples in phase two The idea is simple The tuples from phase one the cheap tuples will be provided by the incomplete modi just discussed In mode 4 mode 1 is used and in mode 5 mode 2 The remainin
177. eset Vv racer read file nrql user guide examples family no signature racer gt OKAY define concrete domain attribute name type string gt OKAY constrained betty betty age age gt OKAY constrained alice alice age age gt OKAY constrained betty betty name name gt OKAY constrained alice alice name name gt OKAY Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 175 N constraints string betty name Betty OKAY Vv constraints string alice name Alice gt OKAY constraints betty age 50 gt OKAY constraints alice age 80 gt OKAY abox consistent gt t Let us enable the data substrate mirroring enable data substrate mirroring gt okay data substrate mirroring enabled describe substrate gt ERROR NRQL ERROR Can t find mirror data substrate of ABox default TRY TO CALL PREPARE NRQL ENGINE FIRST However the data substrate is not created before the first query has to be answered N retrieve x x top x alice x betty x charles x eve x doris v N describe substrate type mirror data substrate abox smith family no of nodes 16 no of edges 24 v We can already see that the mirror substrate contains 16 nodes as well as 24 edges A data substrate node is created in the mirror e for each AB
178. esigned to compute the n dimensional complement set of its n dimensional input argument set Thus neg preserves the arity of its argument However in some cases the dimensionality of the argument shall be reduced before neg is applied This is what project to achieves For example suppose we are looking for the woman which have a male child this is easy full reset gt okay full reset racer read file nrql user guide examples family racer gt OKAY N retrieve x and x woman x y has child y man x alice v But how can we retrieve the complement set i e the woman which do not have a male child In a first attempt we try retrieve x neg and x woman x y has child y man gt x charles x eve x doris x betty x alice So why is alice included in this set If we also put y in the query head it becomes clear what happened retrieve x y neg and x woman x y has child y man gt x charles y eve x charles y doris Racer Systems GmbH amp Co KG http www racer systems com 144 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL x charles y charles x charles y betty x charles y alice x eve y eve x eve y doris x eve y charles x eve y betty x eve y alice x doris y eve x doris y doris x doris y charles
179. esult in assertions such as instance betty woman and instance betty human etc Racer Systems GmbH amp Co KG http www racer systems com 214 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL But carefully For big ABoxes containing many different concept expressions this process might take a long time since each of this concept expressions must be classified into the TBox In order to enable this mode use set nrql mode 2 Complete RacerPro ABox reasoning Mode 3 We don t need to discuss this mode since it is the default mode and has been discussed all the time in this manual In order to enable this mode use set nrql mode 3 See below for an explanation of set nrql mode 4 and set nrql mode 5 If you use set nrql mode 6 then this mode behaves like set nrql mode 3 Thus mode 6 is a complete mode However mode 6 might be faster than mode 3 See below for an explanation Please note that the incomplete modi will only achieve a certain degree of completeness if you restrict yourself to concept query atoms which use only concept names instead of arbitrary concept expressions If you insist on using complex concepts in the concept query atoms of your nRQL queries then the complete nRQL modi will be needed 6 2 5 2 An Example Demonstrating the Different nRQL Modi The following log demonstrates the different nRQL modi and the degrees of completeness they achieve full reset gt okay full r
180. eveals implicit subsumption relationships Im plicit subsumption relationships are indicated in the Prot g window Inferred Hierarchy see Figure 2 17 For instance an uncle is also a brother Prot g nicely summarizes classification results in the list of changed superclasses Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO 37 Kaki family 1 Prot g 3 1 beta file Users rm Desktop family 1 pprj OWL Files owl or rdf File Edit Project OWL Code Window Tools Help DCH B8G wad 49 GET oe D EE db lt Klprot g OWLClasses i Properties Forms Individuals Metadata INSTANCE BROWSER INDIVIDUAL EDITOR FT For Project family 1 For Class family MOTHER For Individual alice instance of family MOTHER Class Hierarchy A Assented Inferred Name Sameas DifferentFrom O Annotations E bf n owl Thing NAME 4axX alice o Property Value Lang family FEMALE alice family HUMAN 2 betty rdfs comment P family MALE gt family mAN Y family PERSON v familyPARENT family FATHER gt familyMOTHER gt family WOMAN famityHAS GENDER ti Fe H hy 6o Name Prefix Conditions O AUNT family family WOMAN n
181. expressed by GCIs There are also separate forms for concept axioms with just concept names on their left hand sides These concept axioms implement special kinds of GCIs and concept equations But concept names are only a special kind of concept terms so these forms are just syntactic sugar They are added to the RacerPro system for historical reasons and for compatibility with KRSS These concept axioms are Primitive concept axioms state the subsumption relation between a concept name and a concept term DL notation CN E C RacerPro syntax define primitive concept CN C Concept definitions state the equality between a concept name and a concept term DL notation CN C RacerPro syntax define concept CN C Concept axioms may be cyclic in RacerPro There may also be forward references to con cepts which will be introduced with define concept or define primitive concept in subsequent axioms The terminology of a RacerPro T box may also contain several axioms for a single concept So if a second axiom about the same concept is given it is added and does not overwrite the first axiom 3 4 Role Declarations In contrast to concept axioms role declarations are unique in RacerPro There exists just one declaration per role name in a knowledge base If a second declaration for a role is given an error is signaled If no signature is specified undeclared roles are assumed to be neither a feature nor a transitive role and th
182. ey do not have any superroles The set of all roles R includes the set of features F and the set of transitive roles RT The sets F and R are disjoint All roles in a T box may also be arranged in a role hierarchy The inverse of a role name RN can be either explicitly declared via the keyword inverse e g see the description of define primitive role or referred to as inv RN Features also called attributes restrict a role to be a functional role e g each individual can only have up to one filler for this role Transitive Roles are transitively closed roles If two pairs of individuals JN and IN and IN and IN3 are related via a transitive role R then JN and IN3 are also related via R Racer Systems GmbH amp Co KG http www racer systems com 3 5 CONCRETE DOMAINS 61 Role Hierarchies define super and subrole relationships between roles If R is a super role of Rg then for all pairs of individuals between which R holds R must hold too In the current implementation the specified superrole relations may not be cyclic If a role has a superrole its properties are not in every case inherited by the subrole The properties of a declared role induced by its superrole are shown in Figure 3 4 The table should be read as follows For example if a role RN is declared as a simple role and it has a feature RN as a superrole then RN will be a feature itself Superrole RNa RIRE F Subrole RN R RIR F de
183. f reals as concrete domain signature atomic concepts teenager roles attributes integer age real temperature celsius real temperature fahrenheit equivalent teenager and human min age 16 equivalent old teenager and human min age 18 equivalent human with fever and human gt temperature celsius 38 5 equivalent seriously ill human and human gt temperature celsius 42 0 Obviously RacerPro determines that the concept seriously ill human is subsumed by human with fever For the reals RacerPro supports linear equations and inequations Racer Systems GmbH amp Co KG http www racer systems com 3 5 CONCRETE DOMAINS 63 Thus we could add the following statement to the knowledge base in order to make sure the relations between the two attributes temperature fahrenheit and temperature celsius is properly represented implies top temperature fahrenheit 1 8 temperature celsius 32 If a concept seriously ill human 1 is defined as equivalent seriously ill human 1 and human gt temperature fahrenheit 107 6 RacerPro recognizes the subsumption relationship with human with fever and the syn onym relationship with seriously ill human In an A box it is possible to set up constraints between individuals This is illustrated with the following extended A box signature atomic concepts teenager roles attributes individuals eve do
184. f the complete answer to a query compared to the answer which would be computed using the complete mode However the incomplete modi can be much more performant and thus achieve better scalability for large ABoxes Also note that the incomplete modi will be complete for simple KBs KBs which do not required the full expressivity of ALCOHIR D7 or OWL The advantages of the incomplete and complete modi can be combined in nRQL The so called two phase processing modi are complete and take care that query process ing works in two phases In phase one the so called cheap tuples are computed and delivered to the client After there are no more cheap tuples the client user can be notified about the upcoming phase transition to phase two nRQL can be advised to deliver a so called warning token before phase two starts The client user can then decide to retrieve these additional expensive tuples or not The remaining tuples from phase two are also called expensive tuples unlike for the tuples from phase one full ABox reasoning is required for their computation e The runtime resources are configurable The API allows to set an upper bound on the number of answer tuples computed a timeout can be specified and the incomplete modi are available Moreover permutations of answer tuples can be excluded Inter nally the engine uses a thread pool The minimum maximum number of threads in this pool can be specified note that a query has to w
185. family HAS SIBLING family PARENT je AllDifferent owl owl Thing E ie AllValuesFromRestriction owl owl Restriction FAO Alt rdfs rdfs Container AnnotationProperty owl rdf Property la AnonymousClass owl OWL CLASS ge BROTHER family family MAN n 3 family HAS SIBLING family PERSON le Bag rdfs rdfs Container g Classification Results Subclasses of owl Thing Figure 2 18 The individuals of the family example Properties of the selected individual alice are displayed After the tab Individuals is selected Prot g displays the individuals of the family knowledge base together with their properties or relationships This is shown here in Figure 2 18 Racer Systems GmbH amp Co KG http www racer systems com 38 CHAPTER 2 USING RACERPRO 8 e ea family 1 Prot g 3 1 beta file Users rm Desktop family 1 pprj OWL Files owl or rdf File Edit Project OWL Code Window Tools Help DEG B6 ad 9 BE OS E OB ap lt lt prot g owLClasses m Properties Forms Individuals Metadata CLASS BROWSER INSTANCE BROWSER INDIVIDUAL EDITOR PORT For Project family 1 For Class family MOTHER For Individual alice instance of family MOTHER Class Hierarchy A Asserted Inferred Name SameAs DifferentFrom J Annotations Ci em owl Thing NAME vox alice le Property Value Lang family FEMALE rs alice family HUM
186. fillers http www owl ontologies com unnamed owl age x 50 Please note that it is possible to specify complex retrieval conditions on the fillers of such datatype properties as the concept query atom x gt http www owl ontologies com unnamed owl age 30 demonstrates The retrieval predicates can be rather complex as the concept query atom x some age or and gt 40 lt 60 80 demonstrates Note that this concept is only permitted as a query concept in a concept query atom but not elsewhere in a RacerPro KB In order to specify retrieval conditions on boolean datatype properties it is important to know how to specify TRUE and FALSE In owl ex4 owl we have added a datatype property married of type boolean to owl ex3 owl full reset gt okay full reset N owl read file nrql user guide examples owl ex4 0w1 home mi wessel nrql user guide examples owl ex4 owl Vv retrieve x datatype fillers married x x boolean married T x http www owl ontologies com unnamed owl alice datatype fillers v http www owl ontologies com unnamed owl married x T retrieve x datatype fillers married x x boolean married F x http www owl ontologies com unnamed owl betty datatype fillers v http www owl ontologies com unnamed owl married x F 6 1 2 3 Retrieving Told Fillers of OWL Annotation Properties In OWL
187. g expensive tuples delivered in phase 2 will be provided by the complete mode 3 Thus all two phase processing modi are complete Note that RacerPro s basic ABox retrieval functions are only called in phase two 6 2 5 3 1 The Warning Token nRQL can be advised to deliver a so called phase two starts warning token before phase two starts informing the application or user that the next call to get next tuple will invoke RacerPro s ABox retrieval functions However delivery of this warning token is optional Please refer to the Reference Manual for more details 6 2 5 3 2 Mode 4 and 5 Cannot be Used for Queries with NEG Please note that only queries that do not make use of the NAF negation neg can take advantage of the two phase query processing modi because the complement set constructed by neg will be too big if neg is applied to a set which contains less tuples than the complete answer Thus the complement will contain wrong incorrect tuples If nRQL is in mode 4 5 or 6 and a containing neg is posed then mode 3 is used Please note that mode 3 can be used in set at a time as well as tuple at a time mode Racer Systems GmbH amp Co KG http www racer systems com 216 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 2 5 3 3 Two Phase Modi Example Session The following log demonstrates the nRQL modi 4 and 6 Let us first set up the small example KB full reset gt okay full reset implies mother woman
188. ge constraints equal age of betty 20 related betty doris has child related betty eve has child related betty charles has sibling closing the role has sibling for charles instance betty at most 1 has sibling Racer Systems GmbH amp Co KG http www racer systems com 245 Charles is the brother of Betty and only Betty instance charles brother Charles s age is 39 constrained charles age of charles has age constrained charles temperature of charles temperature fahrenheit constraints equal age of charles 39 temperature of charles 107 6 related charles betty has sibling closing the role has sibling for charles instance charles at most 1 has sibling Doris has the sister Eve related doris eve has sister instance doris at most 1 has sibling Doris s age is 2 constrained doris age of doris has age constrained doris temperature of doris temperature celsius constraints equal age of doris 2 temperature of doris 38 6 3 Eve has the sister Doris related eve doris has sister instance eve at most 1 has sibling Eve s age is 1 constrained eve age of eve has age constraints equal age of eve 1 Racer Systems GmbH amp Co KG http www racer systems com 246 APPENDIX B A KNOWLEDGE BASE WITH CONCRETE DOMAINS 333 some T box queries 33 are all uncles brothers concept subsumes brother uncle 33 get all super concept
189. gt OKAY instance i top gt OKAY instance j c gt OKAY instance k c gt OKAY process tuple at a time gt o0kay processing tuple at a time enable lazy tuple computation gt okay lazy mode enabled retrieve x y x y not r gt query 1 running get next tuple last gt x k y k get next tuple last gt warning kb has changed active queries gt query 1 waiting queries gt query 1 running queries gt NIL retrieve x x c NRQL WARNING DENIED DUE TO DEADLOCK PREVENTION THE FOLLOWING QUERIES WILL NOT TERMINATE AUTOMATICALLY SINCE THEY HAVE BEEN STARTED IN LAZY INCREMENTAL MODE query 1 gt denied due to deadlock prevention get answer last Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 219 gt x k y k x x x x x retrieve gt query 2 k y j j y K j Cy j i y k i y j x x c running get answer last v x j x k Racer Systems GmbH amp Co KG http www racer systems com 220 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 2 7 Reasoning with Queries The services described in this Section should be considered as non essential add ons and are still experimental 6 2 7 1 Reporting Inconsistent Queries If you ar
190. gt NIL Please note that certain names are reserved for nRQL and thus cannot be used for individ uals see Section 6 1 9 For example in the query retrieve nil and c and is not recognized as an individual name and thus the query is not recognized as a concept query atom A syntax error is raised instead Moreover nRQL establishes a special semantics for individuals which are used in the head of a query Consider the query retrieve betty betty woman One possible way to answer this query would be to simply return TRUE as if the head was empty However nRQL returns retrieve betty betty woman gt betty betty Racer Systems GmbH amp Co KG http www racer systems com 96 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL The variable betty is called the representative variable for betty Internally nRQL rewrites queries by replacing individuals with representative variables describe query last gt query XX accurate processed retrieve betty and same as betty betty betty woman abox smith family 6 1 1 1 4 Querying OWL and RDF S Documents with Concept Query Atoms OWL and RDF S KBs documents can be queried with concept query atoms as well RacerPro represents OWL RDF S individuals and their interrelationships in the ABox OWL RDF S classes as concepts and properties as roles in the TBox Consider the following OWL document nrgql user guide examples
191. hical interfaces or client applications can be executed on you personal computer or via wireless connections on your portable computer Your license file is available from your personal download page the file is called license racerlicense Put this license file in your home directory Do not edit or delete this file If the installation process is executed successfully the RacerPro application comes with an additional license file named racerpro lic in the same directory as the RacerPro executable Do not remove or edit this file either 1 3 4 System Requirements RacerPro is available for all major operating systems in 32bit and 64bit modes Linux Mac OS X Solaris 2 Windows For larger knowledge bases in industrial projects we recommend at least 1 GB of main memory and a 1GHz processor For large ontologies gt 100000 concept and role names we recommend 64bit computer architectures 1 4 Acknowledgments RacerPro is based on scientific results presented in publications about theoretical and prac tical work on KRIS FaCT and other description logic systems See for instance 3 5 7 8 14 15 10 11 12 We would like to thank the team at Franz Inc http www franz com for their col laboration and for the support in making RacerPro one of the fastest and most expres sive commercial OWL RDF and description logic reasoning systems The graphical user interface RacerPorter is built with the development environm
192. i OWL Files owl or rdf File Edit Project OWL Code Window Tools Help De of ea 5 Run ontology tests eo Show TODO list OWLClasses Test settings SUBCLASS RELATIO Check consistency For Project famil Classify taxonomy Compute inferred types d Reasoner inspector Determine Convert OWL Sublanguage EB lt a gt Hi prot g data instance of owl Class Asserted Hierarchy ferentFrom lt owl Thing family FEMALE family HUMAN EE protege OWL Syntax famiy MALE Display prose as tool tip of OWL expressions gt family MAN gt familyPERSO Preferences gt fami WOMAN Dublin Core metadata Define import policy _J Annotatio OWL Plugin Version Info NECESSARY amp SUFFICIENT NECESSARY Figure 2 12 First step for declaring RacerPro as the reasoner used by Prot g In order to declare RacerPro as the standard reasoner used by Prot g select the Preferences menu in Prot g see Figure 2 12 Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO 33 File Edit Proiect OWL Code Window Tools Help O Den f OWLClass SUBCLASS RE I For Project Asserted Hierg family FE family Ht family M gt 8 family M gt family PE gt O family
193. ication from the krss group of the arpa knowledge sharing ef fort Technical report Bell Labs 1993 Available as http www db research bell labs com user pfps papers krss spec ps Racer Systems GmbH amp Co KG http www racer systems com Index Prot g 31 RacerEditor 27 RacerMaster 6 RacerPlus 5 RacerPorter 5 9 25 RacerPro 5 AND query 119 HAS KNOWN SUCCESSOR query atom 109 NEG query 124 PROJECT TO operator 130 SAME AS query atom 107 UNION query 120 ABox augmentation 139 ABox modification 139 active domain semantics 93 APIs 20 assertion 65 associated ABox individual 150 associated Abox individual 150 associated substrate node 150 binary query atom 92 body projection operator 130 boolean query 130 bug reports 24 cheap query 182 cheap rule 182 cheap tuple 182 closed world assumption 73 complex query 118 complex TBox query 144 concept axioms 59 concept definition 60 concept equation 59 concept query atom 93 Racer Systems GmbH amp Co KG http www racer systems com 262 INDEX concept term 56 concrete domain restriction 59 concrete domain value query 110 concrete domains 61 concurrent query processing 173 conjunction of roles 61 conjunctive query 119 consistency of qualitative spatial networks 161 constraint query atom 102 creating individuals with a rule 140 data representation 149 data substrate 149 data subst
194. ieve x union x c x not c gt x i x j retrieve x x or c not c gt x K x j x i 6 1 4 3 NEG The Negation As Failure Constructor A NEG constructor is provided which provides Negation as Failure short NAF negation NAF is especially useful for measuring the completeness of the modeling in a KB Many users are probably familiar with logic programming language the PROLOG which also offers NAF NAF is quite different from classical true negation The neg operator can be applied to an arbitrary query body 6 1 4 3 1 Negation as Failure for Concept Query Atoms NAF is quite different from true negation Consider the following example full reset gt okay full reset racer read file nrql user guide examples family racer gt OKAY retrieve x x grandmother x alice v N retrieve x neg x grandmother x eve x doris x charles x betty v RacerPro can prove that alice is an instance of grandmother Thus for the other present individual it cannot prove membership in grandmother In order to retrieve the individu als for which RacerPro cannot prove membership the neg operator is provided We can Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 131 simply retrieve these individuals with the NAF negated concept query atom neg x grandmothe
195. ifferent querying modi The two major modi are the set at a time mode and the tuple at a time mode 6 2 1 1 The Set at a Time Mode This is the mode we have used so far in our examples You are already familiar with the behavior of the nRQL API in this mode describe query processing mode gt creating substrates of type racer dummy substrate using prefix for injective variables check abox consistency same as is syntactic same as query optimization enabled optimizer uses cardinality heuristics automatically adding rule consequences warnings complete mode mode 3 set at a time mode deliver kb has changed warning tokens retrieve x x woman gt x alice x betty x eve x doris Note the set at a time mode token in the result of describe query processing mode In this mode a call to retrieve returns the whole query answer at once in one bunch Most users and or client applications will be happy with this mode 6 2 1 2 The Tuple at a Time Modi Incremental Iterator Based Query Pro cessing Sometimes it is preferable to compute and retrieve the answer tuples in an incremental iterator based fashion The user or application requesting the tuples can look at the tuples already retrieved and decide whether yet another tuple is needed or not Racer Systems GmbH amp Co KG http www racer systems com 204 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL The nRQ
196. iller of the has child role of x such that can be formulated anyway by ex ploiting the expressive power of the concept expressions used in concept query atoms Simply use concept query atoms such as x some has child nRQL offers a variety of query atoms the most important ones are concept query atoms role query atoms constraint query atoms and SAME AS query atoms Complex queries query bodies are built from query atoms using the operators and union neg project to Negation as failure NAF as well as true classical negation is available NAF is especially useful for measuring the degree of completeness of a modeling of the domain of discourse in the KB Special support for querying the concrete domain part of an ABox nRQL allows for the formulation of complex retrieval conditions on concrete domain attribute fillers of ABox individuals by means of complex concrete domain predicates Concrete domain predicates can appear in concept query atoms as well as in constraint query atoms For example we can retrieve those individuals whose fillers of the concrete domain attribute has age satisfy the complex concrete domain predicate gt 18 using a concept query atom x min age 18 Moreover in case satisfying concrete domain values are known in the ABox they have been told to the system these known told values if existing can be returned as part of the query answer Thus in case the concrete age being a fille
197. ine is a multi threaded query answering engine which allows you to run several queries concurrently in parallel Thus is is possible to submit a number of calls to retrieve to the engine and then request tuples from these queries in a random order retrieve x x man gt query 4 running retrieve x x uncle gt query 5 running get next tuple query 4 gt x charles get next tuple query 5 gt x charles get next tuple query 4 gt exhausted get next tuple query 5 gt exhausted 6 2 1 2 2 Lazy and Eager Tuple at a Time Modi The incremental mode comes in two forms lazy and eager e In the so called lazy incremental mode the next tuple of a query is not computed before it is actually requested by the user or application unless get next tuple is called The thread computing the tuples which is called the query answering thread in the following is put to sleep and get next tuple re awakes it e In the so called eager incremental mode the query answering thread is not put to sleep Instead it continues to pre compute tuples even if the application has not yet requested these future tuples These tuples are put into a queue for future requests The API behaves identically in both modi In any case the query answering thread dies after all tuples have been computed or if the query was aborted or a timeout was reached or the upper bound on the numbe
198. ing 161 spatial reasoning substrate 161 spatio thematic query 162 stored defined query 201 stored QBox 201 stored substrate 201 substrate 149 substrate layer 149 substrate persistency 201 substrate query 151 SWRL 44 system requirements 6 TBox query 144 told value query 110 transitive role 60 tuple at a time mode 172 two phase query processing 182 unary query atom 92 unique name assumption 65 73 value restriction 57 variable 92 warning token 182 Racer Systems GmbH amp Co KG http www racer systems com
199. ion ready queries returns this list The function query ready p is used to check whether a query is on this list The queries on this list are called ready queries or prepared queries The queries on this list have been parsed and compiled prepared but have not been started yet i e no query answering thread have been assigned to them The API function execute query can be applied to queries on this list In order to put a query to this list without also starting its query answering thread use the API function macro prepare abox query instead of retrieve 3 The list of active queries These queries have been started and there is an ac tive query answering thread associated with them The corresponding API function active queries returns this list The function query active p is used to check whether a query is on this list According to the mode in which the query had been started this list is further par tioned into the following two list Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 207 a The list of queries which are currently running The corresponding API function running queries returns this list The function query running p is used to check whether a query is on this list A query which appears on this list is actively consuming CPU time its thread is not sleeping 2 The list of queries which are currently waiting sleeping The corresponding API
200. ion with negation as failure this operator can be used to express closed world universal quan tification which is important for many retrieval tasks nRQL is also a powerful OWL and RDF S query language In contrast to many other RDF S query languages such as Sparql nRQL uses description logic reasoning to retrieve also inferred triples and thus does not work on the syntactic level of triples but on the level of semantic models Like RDF s query languages nRQL allows for the specification of retrieval predicates for data values which are instances of XML Schema datatypes In Sparql these retrieval conditions are specified using the filter construct There are two main options to specify such retrieval conditions 1 The concrete domain constraint checking facilities also apply to datatype prop erties Thus a datatype property can be used as if it were a concrete domain attribute This applies to concept as well as to constraint query atoms For con cept query atoms we have extended the concept expression syntax to support this However the language which can be used for formulating such predicates is limited to the predicates of the different concrete domains offered by RacerPro Moreover certain predicates cannot be supported at all in order to guarantee decidability in the RacerPro concept language 2 However additional and more complex retrieval predicates can be specified with queries posed to the so called data su
201. is y doris doris y betty doris y alice charles y eve charles y doris charles y betty charles y alice betty betty betty betty alice alice alice alice y y y y y y y y eve doris betty alice eve doris betty alice Note that the atom without NAF returns 5 tuples recall that charles cannot be a sister since he is male retrieve x y x y not has sister gt x eve y charles x doris y charles x charles y charles x betty y charles x alice y charles Thus we get our 20 5 25 5 5 pairs again 6 1 4 3 4 Negation as Failure for Constraint Query Atoms Also constraint query atoms can be NAF negated However for simple CD predicates NAF negation and classical negation coincide since the CDs in RacerPro are complete theories thus for two concrete domain objects i 7 either F P i 7 or P i 7 holds This is demonstrated using the CD predicate please note that for each CD predicate a negated predicate is available in this case the negated predicate for is lt gt retrieve x y meg x y constraint age age gt x eve y doris x eve y charles x eve y betty Racer Systems GmbH amp Co KG http www racer systems com 138 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL x eve y alice x doris y eve
202. is constructed The second query is in fact equivalent to retrieve x meg project to x x y has child gt x eve x doris x charles The atom x has known successor has child is therefore just syntactic sugar for the body neg project to x x y has child 6 1 1 5 2 BIND INDIVIDUAL Query Atoms A bind individual lt name gt query atom for some individual lt name gt is just syntactic sugar for same as lt name gt lt name gt where lt name gt is the representative variable for lt name gt These atoms are provided for backward compatibility with older versions of nRQL only 6 1 2 Query Head Projection Operators So far we have only used simple objects variables and individuals in the head of a query Additionally so called head projection operators are admitted as head entries as well A head projection operator is simply a function operator which is applied to the current binding of the variable the current individual and the operator result is included in the answer tuple There are various predefined operators available nRQL also offers a simple termination safe expression language called MiniLisp Users can define their own projection operators as lambda head operators Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 111 6 1 2 1 Retrieving Told Values from the Concrete Domain We already mentioned that variables can only be
203. iscuss each type of atom 6 1 1 1 Concept Query Atoms Concept query atoms are unary atoms A concept query atom is used to retrieve the in stances of a concept or an OWL or RDF S class for example the instances of the concept woman in family racer We use RacerPorter to read in that file into RacerPro note that denotes the input prompt and gt the RacerPro result Please use the Shell Tab of Racer Porter and enter the following commands or use the button Load KB which will send the same command to read in the family racer KB full reset gt okay full reset racer read file nrql user guide examples family racer gt OKAY Now we retrieve the instances of class woman using a query whose body consists of a single concept query atom x woman retrieve x x woman gt x betty x eve x doris x alice The format of the query answer is specified by the query head see below In this case the head is given as x Note that this query has the same semantics as concept instances woman but the result set is not returned as a single set but as a set of so called binding lists or componentwise named tuples Note that a set of tuples is returned We emphasize set here because this implies that there is no guarantee on the order of the tuples in this set However the result tuples itself are ordered of course Moreover complex concept term
204. ivalence classes use individual synonym equivalence classes gt okay using individual equivalence classes retrieve x x good man gt x santa claus Using a so called head projection operator see below the synonyms can be delivered as well retrieve x individual synonyms x x good man dont show lambdas p t gt x santa claus santa claus weihnachtsmann In this mode the same as atom no longer denotes name equality but semantic equal ity Thus same as x y becomes equivalent to x y nrql equal role and the nrql equal role shouldn t be used Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 109 retrieve x same as x weihnachtsmann gt x santa claus Finally we want to mention that the same as assertion is not the only way how to two individuals can be forced to become synonyms Also features functional roles or so called number restrictions can cause individuals to become synonyms for example full reset okay full reset v define primitive role f feature t gt OKAY related i j f gt OKAY related i k f gt OKAY related 1 m r gt 0KAY related 1 n r gt OKAY instance 1 at most 1 r gt OKAY retrieve x y and x y nrql equal role neg same as x y gt x n y m x m y n x k Py j CPx j Cry k Plea
205. ive ABox re trieval functions are avoided whenever possible However since phase one is incomplete it may be the case that no tuples can be computed in phase one at all thus leaving all the work for phase two Nevertheless it is worth trying to use mode 6 whenever you encounter performance problems in mode 3 6 2 6 Automatic Deadlock Prevention Certain types of queries or rules must temporary change the queried ABox in order to be answered i e ABox assertions must be added This is for example the case if classical negated roles are used in role query atoms If more than one query or rule is currently active which reference the same ABox then these different queries most be isolated from one another nRQL uses locking techniques to achieve this isolation The queried ABox is thus locked so other queries cannot access this ABox as long as the query is active This means that nRQL may not permit the execution of other queries on that ABox The query has to wait until all active queries querying that ABox have been terminated However there is a problem In lazy mode these queries may not terminate automatically If such a situation is encountered nRQL will deny starting yet another query and print out a warning message full reset Racer Systems GmbH amp Co KG http www racer systems com 218 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL gt okay full reset define primitive role r range not c
206. ividuals are declared and published i e the documents are potential results of subscription based instance retrieval queries state instance document 1 article publish document 1 instance document 2 book constrained document 2 isbn 2 isbn constraints equal isbn 2 2234567 publish document 2 instance document 3 book constrained document 3 isbn 3 isbn constraints equal isbn 3 3234567 publish document 3 instance document 4 book constrained document 4 isbn 4 isbn constraints equal isbn 4 4234567 publish document 4 instance document 5 computer science book constrained document 5 isbn 5 isbn constraints equal isbn 5 5234567 publish document 5 Racer Systems GmbH amp Co KG http www racer systems com 5 3 THE PUBLISH SUBSCRIBE MECHANISM 81 Now we assume that a client subscribes to a certain instance retrieval query state subscribe client 1 book The answer returned by RacerPro is the following CLIENT 1 DOCUMENT 2 CLIENT 1 DOCUMENT 3 CLIENT 1 DOCUMENT 4 CLIENT 1 DOCUMENT 5 RacerPro returns a list of pairs each of which consists of a subscriber name and an individual name In this case four documents are found to be instances of the query concept subscribed und the name client 1 An application receiving this message from RacerPro as a return result can then decide how to inform the client appropriately In future r
207. jection list must be a subset of the set of objects mentioned in the query body Please look up the EBNF specification to learn more about the compositional syntax of nRQL see Section 6 1 9 We will now present and discuss each body constructor 6 1 4 1 The AND Constructor Conjunctive Queries Suppose we want to retrieve all woman having a male child from family racer This is a classic grounded conjunctive query retrieve x y and x woman x y has child y man gt x alice y charles Here is another example query which searches for children having a common mother retrieve mother childi child2 and Racer Systems GmbH amp Co KG http www racer systems com 128 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL childi human child2 human neg same as child1 child2 mother childi has child mother child2 has child gt mother betty childi doris child2 eve mother betty child1 eve child2 doris mother alice child1 betty child2 charles mother alice childi charles child2 betty Note that the atom neg same as childi child2 prevents that child1 and child2 are bound to the same individual 6 1 4 2 The UNION Constructor nRQL also offers a union constructor also or can be used retrieve x union x woman x man gt x alice x betty x eve x doris
208. knowledge base see the file family 2 racer in the examples folder in knowledge base family define primitive role descendants transitive t define primitive role children parents descendants implies and male female bottom equivalent man and male human equivalent woman and female human equivalent parent at least 1 children equivalent grandparent some children parent equivalent mother and woman parent equivalent father and man parent implies some descendants human human implies human all descendants human equivalent father having only male children and father human all children male equivalent father having only sons and man at least 1 children all children man equivalent grandpa and male some children and parent human equivalent great grandpa and male some children some children and parent human instance john male instance mary female related john james children related mary james children instance james and human male Racer Systems GmbH amp Co KG http www racer systems com 240 APPENDIX A ANOTHER FAMILY KNOWLEDGE BASE instance john at most 1 children individual direct types john individual direct types mary individual direct types james Racer Systems GmbH amp Co KG http www racer systems com Appendix B A Knowledge Base with Concrete Domains In this section we present another family know
209. l 2 Ready A Figure 2 24 RacerPorter used to query an ontology edited and verified with Prot g Te 00 RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles indiv Asse Quer Rules TBox Ihttp dI Kr org dig kb 471611 ABox Ihttp dl kr org dig kb 471611 Concept Role Individual Query Rule Simplify Names Id State Cheap Next Query QUERY 1 ACCURATE PROCESSED yes no RETRIEVE X CX Country ABOX http dl kr org dig kb 47161 E J gt all O Ready O Running O waiting O Cheap w O Expensive w O Processed O ace O Inace Status GetNext Tuple Answer Delete Abort Execute Re Execute Re Prepare RacerPro Log a 2 RETRIEVE X X Country m 2 gt CCC X IFrancel CC X IGermanyl CC X lAmerical CCX IItaly1 CC X IEnglandl O X A Figure 2 25 Using RacerPorter queries can be inspected as objects Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO 43 2 5 4 TopBraidComposer TopBraidComposer www topbraid com is a well supported commercial ontology editor see Figure 2 26 Currently TopBraid Composer can be used via the DIG interface In the near future a more efficient dynamic link library will support a more seamless integration of RacerPro and TopBraid Composer G 5 n n gt
210. le has gender is a feature in family racer retrieve person gender person gender has gender gt NIL So the answer set is empty Why is this The reason is that there are no explicit gender fillers of the has gender feature in that ABox Thus gender cannot be bound to an ABox individual There are two solutions Either new gender individuals could be added with a rule see below or a concept query atom can be used to retrieve for example those instances with female gender retrieve x x some has gender female gt x betty x eve x doris x alice 6 1 1 2 1 Role Terms in Role Query Atoms Whereas arbitrary concept terms expressions not only atomic concepts also called concept names are admissible in concept query atoms role terms are admissible in role query atoms not only atomic roles also called role names nRQL admits inverse and negated roles retrieve child of betty child of betty betty inv has child gt child of betty eve child of betty doris Please note that inv has child is a role term The second kind of role term constructor is given by the not operator which is used to construct a negated role Let us discuss negated roles Racer Systems GmbH amp Co KG http www racer systems com 100 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 1 1 2 2 Classical Negated Roles in nRQL Availability of classical negated roles is
211. le traffic lights one of red green The following example demonstrates the use of A boxes and the string concrete domain to provide for a formal model of a crossing with simple traffic lights in knowledge base traffic lights define concrete domain attribute color type string define concept colorful object or string color red string color green define concept traffic light and a color colorful object instance traffic light 1 traffic light instance traffic light 2 traffic light instance traffic light 3 traffic light instance traffic light 4 traffic light constrained traffic light 1 color traffic light 1 color constrained traffic light 2 color traffic light 2 color constrained traffic light 3 color traffic light 3 color constrained traffic light 4 color traffic light 4 color constraints string color traffic light 1 color traffic light 3 constraints string color traffic light 2 color traffic light 4 constraints string lt gt color traffic light 1 color traffic light 2 constraints string lt gt color traffic light 2 red constraint entailed string color traffic light 2 green constraint entailed string color traffic light 4 green constraint entailed string color traffic light 1 red constraint entailed string color traffic light 3 red The four queries at the end all return t for true although only indefinite information is
212. le interface fully supports the DIG standard for queries see above Currently the file interface of RacerPro supports only queries given in KRSS syntax How ever DIG based queries 4 can be specified indirectly with the file interface as well Let us assume a DIG knowledge base is given in the file kb xml and corresponding DIG queries are specified in the file queries xml In order to submit this file to RacerPro just create a file q racer say with contents dig read file queries xml and start RacerPro as follows RacerPro dig kb xml q q racer Under windows you might want to suppress the console window RacerPro c dig kb xml q q racer Note the use of the option dig for specifying that the input knowledge base is in DIG syn tax Since the file extension for the knowledge base is xml the option f would assume the Racer Systems GmbH amp Co KG http www racer systems com 20 CHAPTER 2 USING RACERPRO older XML syntax for knowledge bases see above If the query file has the extension xm1 RacerPro assumes DIG syntax For older programs this kind of backward compatibility is needed RacerPro opens one TCP port for the DIG protocol and one for RacerPro s native protocol The DIG protocol is an HTTP based XML protocol and thus is based on plain text The native protocol is proprietary to Racer but in KRSS syntax and is documented in the Reference Manual Since both protocols are based on plain text
213. le to locally minimize computation time lazy inference mode For instance in order to answer a subsumption query w r t a T box it is not necessary to classify the T box However once a T box is classified answering subsumption queries for atomic concepts is just a lookup Furthermore asking whether there exists an atomic concept in a T box that is inconsistent tbox coherent p does not require the T box to be classified either In the lazy mode of inference the default RacerPro avoids computations that are not required concerning the current query In some situations however in order to globally minimize processing time it might be better to just classify a T box before answering a query eager inference mode A similar strategy is applied if the computation of the direct types of individuals is requested RacerPro requires as precondition that the corresponding T box has to be classified If the lazy inference mode is enabled only the individuals involved in a direct types query are realized We recommend that T boxes and A boxes should be kept in separate files If an A box is revised by reloading or reevaluating a file there is no need to recompute anything for the T box However if the T box is placed in the same file reevaluating a file presumably causes the T box to be reinitialized and the axioms to be declared again Thus in order to answer an A box query recomputations concerning the T box might be necessary So if
214. ledge base with concrete domains see the file family 3 racer in the examples folder in knowledge base family smith family signature atomic concepts human female male woman man parent mother father mother having only female children mother having only daughters mother with children mother with siblings mother having only sisters grandpa great grandpa grandma great grandma aunt uncle sister brother sibling young parent normal parent old parent child teenager teenage mother young human adult human old human young child human with fever seriously ill human human with high fever roles has descendant domain human range human transitive t has child domain parent range child parent has descendant has sibling domain sibling range sibling has sister range sister Racer Systems GmbH amp Co KG http www racer systems com 242 APPENDIX B A KNOWLEDGE BASE WITH CONCRETE DOMAINS parent has sibling has brother range brother parent has sibling features has gender range or female male attributes integer has age real temperature fahrenheit real temperature celsius individuals alice betty charles doris eve objects age of alice age of betty age of charles age of doris age of eve temperature of doris temperature of charles Racer Systems GmbH amp Co KG http www racer systems com 243 3 the concepts disjoint female male human implies human an
215. les alo apa Name Expression Rule 1 hasMother x 7y A hasBrother y z A differentFrom x 7y hasUncle x z A k ocrtane cD OWLClasses Figure 2 27 Snapshot of the famous uncle rule in Prot g Rules can be edited for instance with the graphical tool Prot g see Figure 2 27 Use the menu Project in Prot g and select Configure Then tick the check box SWRLTab in order to activate the SWRL editor in Prot g After the ontology is saved the OWL SWRL specification can be interpreted by applications For this purpose a reasoner such as Racer Pro is required In Figure 2 28 the source code of the uncle rule is shown in the RacerPro Editor The editor also shows some instances MARY has a mother SUE who in turn has a brother JOHN The rule is responsible for asserting that MARY has an uncle who is JOHN With RacerEditor the OWL SWRL file can be sent to RacerPro by selecting a menu item or pressing a key combination Racer Systems GmbH amp Co KG http www racer systems com 2 6 SWRL SEMANTIC WEB RULE LANGUAGE 45 099 RacerEditor EP xm version 1 0 gt A lt rdf RDF xmlns swrlb http www w3 0rg 2003 11 swrlb xmlns swrl http ww w3 0rg 2003 11 swrl xmlns rdf http ww w3 0rg 1999 02 22 rdf syntax ns xmlns xsd http www w3 0rg 2001 XMLSchema xmlns rdfs http www w3 0rg 2000 01 rdf schema xmlns owl http www w3 0rg 2002 07
216. lns xsd http www w3 0rg 2000 10 XMLSchema gt lt rdf RDF gt By default RacerPro prepends the URL of the default namespace to all OWL names starting with the sign If you would like to instruct RacerPro to return abbreviated names i e to remove the prefix again in output it produces start the RacerPro server with the option n Individual names nominals in class declarations introduced with owl oneOf are treated as disjoint atomic concepts This is similar to the behavior of other OWL inference en gines Currently RacerPro can provide only an approximation for true nominals Note that reasoning is sound but still incomplete if owl oneOf is used In RacerPro individuals used in class declarations are also represented in the A box part of the knowledge base They are instances of a concept with the same name An example is appropriate to illustrate the Racer Systems GmbH amp Co KG http www racer systems com 4 7 OWL INTERFACE 79 idea Although the KRSS syntax implemented by RacerPro does not include one of as a concept building operator we use it here for demonstration purposes in knowledge base test implies c some r one of j instance i c Dealing with individuals is done by an approximation such that reasoning is sound but must remain incomplete The following examples demonstrate the effects of the approximation Given this knowledge base asking for the role fillers of r w r t i returns
217. logies com Ontology1159352693 o0owl test y 123 retrieve x x owl object gt x http www owl ontologies com Ontology1159352693 owl test x 123 Racer Systems GmbH amp Co KG http www racer systems com 186 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL x Comment for individual test x This is a test class x 456 retrieve x y x y owl relationship gt x http www owl ontologies com Ontology1159352693 owl test y http www owl ontologies com Ontology1159352693 owl test x http www owl ontologies com Ontology1159352693 owl test y 123 x http www owl ontologies com Ontology1159352693 owl test y Comment for individual test x http www owl ontologies com Ontology1159352693 owl test y 456 x http www owl ontologies com Ontology1159352693 owl class y This is a test class 6 1 8 3 The RCC Substrate Querying RCC Networks Many applications have to represent domain objects that also have spatial characteristics However the applicability of standard DLs in spatial domains is limited To support such applications we are offering a spatial representation layer the RCC substrate The so called Region Connection RCC Calculus provides a well known and widely used set of qualitative spatial relationships QSRs The QSRs offered by RCC are used to describe the relative location of spati
218. lt contains rdf resource germany gt lt continent gt lt rdf RDF gt So the object properties contains adjacent and disjoint are meant to denote qualitative spatial relationships in the OWL KB In RCC8 terminology these QSRs can be expressed as the disjunctive relations ntppi tppi EC and DC If we declare Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 191 these object properties as RCC synonyms then the RCC substrate will recognize and process them accordingly owl read file nrql user guide examples owl rcc owl gt nrql user guide examples owl rcc owl enable rcc substrate mirroring gt okay rcc substrate mirroring enabled rcc synonym contains ntppi tppi gt http www owl ontologies com Ontology1162148702 owl contains ntppi tppi rcc synonym adjacent ec gt http www owl ontologies com Ontology1162148702 owl adjacent ec http www owl ontologies com Ontology1162148702 owl contains ntppi tppi rcc synonym disjoint dc gt http www owl ontologies com Ontology1162148702 owl disjoint dc http www owl ontologies com Ontology1162148702 owl adjacent ec http www owl ontologies com Ontology1162148702 owl contains ntppi tppi These RCC synonyms can be used in queries as well retrieve x y and x y adjacent gt x http www owl ont
219. mainConcept Country lceCream v Pizza Y CheeseyPizza American AmericanHot Cajun Capricciosa Caprina Fiorentina FourSeasons Giardiniera LaReine Margherita Mushroom Napoletana t Parmense ez Metadata uals For Class Capricciosa instance of owl Class Inferred Hierarchy a Ac f Name SameAs DifferentFrom O Annotations a Property Capricciosa 9 DJ rdfs label Value Lang Capricciosa pt rdfs comment en coperti Hi i EG Lj PIP gt O hasBase single PizzaBase O hasTopping a multiple PizzaTop Asserted Conditions NamedPizza WV hasTopping TomatoTopping HamTopping E 3 hasTopping PeperonataTopping 3 hasTopping AnchoviesTopping 3 hasTopping CaperTopping a hasTopping OliveTopping 3 hasTopping HamTopping a hasTopping MozzarellaTopping a hasTopping TomatoTopping ED Disjoints FourSeasons QuattroFormaggi AmericanHot INHERITED B bs Logic View O Properties View ajs American AmericanHot Cajun CajunSpiceTopping ES Capricciosa x Caprina r ale Class I Changed superclasses Added CheeseyPizza InterestingPizza MeatyPizza a Added InterestingPizza CheeseyPizza SpicyPizzaEquivalent MeatyPi
220. member it was soccer or basket ball so he must be a soccer player or a basket ball player Being able to deal with indefinite information of this kind is important when information about data comes in from various sources and in situations where sources are for instance used to exclude certain cases that are possible given the information at hand Description logics are also important if data descriptions are to be queries with respect to varying T boxes or ontologies Note that this is a common scenario in modern information technology infrastructure Due to the rapid pace in technology evolution also the vocabulary used to access data descriptions changes quite frequently New concepts are introduced and can be set into relation to older terminology using description logics or semantic web ontologies There are numerous papers describing how description logic in general and RacerPro in particular can be used to solve application problems see the International Workshops on Description Logics and the workshops on Applications of Description Logics ADL 01 ADL 02 or ADL 03 Without completeness one can summarize that applications come from the following areas Racer Systems GmbH amp Co KG http www racer systems com 1 3 RACER EDITIONS INSTALLATION LICENSES AND SYSTEM REQUIREMENTS 5 e Semantic Web Semantic Grid ontology representation and logic based information retrieval e Electronic Business e g reason
221. ms x x taxonomy node dont show lambdas p t gt x top top top x woman with children woman with children mother x woman woman x bottom bottom bottom gt tbox retrieve x describe x x woman dont show lambdas p t gt x woman woman told primitive definition nil synonyms woman parents top top children woman with children mother To retrieve all subsumer or subsumees use the has descendant and or has ancestor role tbox retrieve y and x woman x y has descendant gt y woman with children y bottom y bottom y mother tbox retrieve y and x woman x y has descendant y taxonomy node y bottom y woman with children v tbox retrieve Racer Systems GmbH amp Co KG http www racer systems com 164 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL y concept synonyms y and x woman x y has descendant y taxonomy node dont show lambdas p t gt y bottom bottom bottom y woman with children woman with children mother Also neg work as expected for example it is easy to retrieve all concepts different from woman tbox retrieve x neg x mother gt x bottom x bottom x woman x top x top Note that also woman with children satisfies mother therefore it is excluded from the answ
222. n x anol y has child y human Po ___ and x anol mother y x anol has parent y man pia Al ___master bottom query in x top 335 ___ a human 335 ___ b woman 333 ___master bottom query gt see output on stdout Racer Systems GmbH amp Co KG http www racer systems com 230 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL show qbox for abox smith family nil 333 QBOX FOR racer dummy substrate FOR ABOX smith family 333 O master top query si le 4 query 9 53 ___5 query 43 query 56 ee ___0 master bottom query 55 6 1 query 1 333 ___2 query 2 SUBQUERY 3 0F query 9 333 ___3 query 5 SUBQUERY 1 OF query 9 335 ___0 master bottom query gt see output on stdout query equivalents last gt query 43 query equivalents query 56 gt query 43 query equivalents query 43 gt query 56 retrieve a a mother gt a betty a alice show qbox for abox smith family nil 333 QBOX FOR racer dummy substrate FOR ABOX smith family 333 O master top query 333 4 query 9 53 ___5 query 43 query 56 oe ___0 master bottom query 55 Nacelequery 1 Hee ___2 query 2 SUBQUERY 3 0F query 9 pa ___3 query 5 SUBQUERY 1 0F query 9 33 ___6 query 88 335 ___0 master bottom query gt see output on stdout Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY P
223. n gt bind individual lt abox query individual gt substitute insert lt query name gt lt query object gt lt data query object gt nil lt query object gt lt data query object gt nil lt query name gt Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 197 u same as mon equal lt query object gt lt query object gt lt query object gt has known successor lt role expression gt lt tbox query atom gt gt true query false query not neg lt tbox query atom gt non inv lt tbox query atom gt my lt query object gt lt concept name gt lt query object gt lt query object gt has child has parent has descendant has ancestor has synonym lt query object gt nil lt role expression gt nil lt query object gt lt role expression gt top lt query object gt bottom lt query object gt bind individual lt abox query individual gt substitute insert lt query name gt lt query object gt lt data query object gt nil lt query object gt lt data query object gt nil lt query name gt wen same as wou equal lt query object gt lt query object gt lt query obje
224. n on the use individual synonym equivalence classes mode This means that synonym individuals will not be recognized as such same as alice grandmother Racer Systems GmbH amp Co KG http www racer systems com 134 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL gt OKAY N retrieve alice alice grandmother alice alice Vv retrieve alice meg alice grandmother alice grandmother alice eve alice doris alice charles alice betty Vv describe query last query 69 accurate processed retrieve alice or not same as alice alice and not alice grandmother top alice abox smith family v So grandmother is returned here even though alice and grandmother are synonyms Due to the standard individual transformation the query alice grandmother is rewrit ten into and alice grandmother same as alice alice Thus neg alice grandmother was consequently rewritten into union alice grandmother neg same as alice alice as shown by the result of describe query last as men tioned not can be used for neg and or for union Since neg same as alice alice holds for the binding alice grandmother this binding is included in the answer What can be done to exclude this binding We can simply replace the automatically added same as atom with a role query atom referencing the nrql equal role al ternati
225. n the XML or OWL RDF interface here in order to directly present the declaration and results of queries in a brief and human readable form Note however that all examples can be processed in a similar way with the XML based interfaces 2 1 Sample Session with RacerPorter The file family racer in the examples folder of the RacerPro distribution con tains the T box and A box introduced in this section The queries are in the file family queries racer In order to run the example just start RacerPro by double clicking the program icon or alternatively type RacerPro as a command in a shell win dow We use the interactive graphical interface to demonstrate the result of queries in this sam ple session However you can also use the RacerPro executable in batch mode If you use the RacerPro executable in batch mode just type RacerPro f family racer q family queries lisp into a shell window in order to see the results under Windows type RacerPro f family racer q family queries lisp See also Section 2 2 for details on how to use RacerPro from a shell You should see something similar to the following 333 Welcome to RacerPro Version 1 9 0 2005 11 21 333 Racer Renamed Abox and Concept Expression Reasoner 333 Supported description logic ALCQHIr D 333 Supported ontology web language subset of OWL DL no so called nominals 333 Copyright C 2004 2005 by Racer Systems GmbH amp Co KG 33 All rights reserved See lice
226. nary Atoms with Individuals A unary atom and its NAF negated variant are always complementary This also holds if the atom refer ences an individual and not a variable N retrieve x x grandmother x alice Vv N retrieve x neg x grandmother x eve x doris x charles x betty v Racer Systems GmbH amp Co KG http www racer systems com 132 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL We have already discussed that individuals in queries are in fact replaced by representative variables e g alice is replaced with alice and a same as alice alice conjunct is added We call this the standard transformation in the following This allows convenient NAF negation but note that in the NAF negated variant alice behaves like a variable retrieve alice alice grandmother gt alice alice describe query last query 7 accurate processed retrieve alice and same as alice alice alice grandmother abox smith family v N retrieve alice neg alice grandmother falice eve alice doris alice charles alice betty v describe query last query 8 accurate processed retrieve alice or v not same as alice alice and not alice grandmother top alice abox smith family However if this behavior is unwanted for neg alice grandmother then simply en sure th
227. nd to contain a lot of annotations and told XML Schema datatype data values thus the query predicates e g search are helpful to retrieve certain resources based on information which is not available for reasoning note that the fillers values of annotation properties are not used for reasoning Note that substrate variables can now be bound to told values of OWL datatype or annota tion fillers So you no longer have to use the head projection operators in order to retrieve told values Moreover retrieval conditions can be specified which couldn t be specified with out the mirror substrate We will use nrql user guide examples owl ex5 0wl which contains all kinds of OWL properties lt xml version 1 0 gt lt rdf RDF xmlns http www owl ontologies com Ontology1159352693 0w1 xmlns rdf http www w3 0rg 1999 02 22 rdf syntax ns xmlns xsd http www w3 org 2001 XMLSchema xmlns rdfs http www w3 0rg 2000 01 rdf schema xmlns owl http www w3 org 2002 07 owl xml base http www owl ontologies com Ontology1159352693 0w1 gt lt owl Ontology rdf about gt lt rdfs comment rdf datatype http www w3 org 2001 XMLSchema string gt Demo Ontology lt rdfs comment gt lt ow1 Ontology gt lt owl Class rdf ID class gt lt rdfs comment rdf datatype http www w3 org 2001 XMLSchema string gt This is a test class lt rdfs comment gt lt owl Class gt lt owl ObjectProperty rdf ID object property
228. ned is also possible to take TBox information into account however in this case the TBox should be classified first in order to get as many answers as possible 6 2 5 1 The nRQL Modi in Detail The nRQL engine offers the following degrees of completeness for query answering which are now discussed in oder of increasing completeness e Told information querying Mode 0 In this mode only the told syntactic information from an ABox the ABox assertions will be used for query answering For example if an ABox contains the assertion instance betty mother then retrieve x x mother will correctly return x betty but the query retrieve x x woman will fail since this information is not explicitly given in the ABox If the ABox contains 2 assertions instance i a and instance i b then both retrieve x x a and retrieve x x b return x i but the query retrieve x x and a b fails as well However retrieve x and x a x b will succeed The mode is therefore severely incomplete but sufficient for data You could also write a nRQL rule to augment the ABox syntacti cally firerule and x a x b instance x and a b In order to achieve better completeness conjunctions in the concept instance asser tions are broken up and made syntactically explicit If an ABox concept assertion such Racer Systems GmbH amp Co KG http www racer systems com 6 2 TH
229. nil Note that OWL names must be enclosed with bars individual fillers file C Ralf Ind Examples ex1 owl i file C Ralf Ind Examples ex1 owl R NIL Asking for the instances of j returns j concept instances file C Ralf Ind Examples ex1 owl j file C Ralf Ind Examples ex1 owl1 j The following knowledge base for the OWL version see file ex2 owl in the examples folder is inconsistent in knowledge base test implies c all r one of j instance i c related i k r Note again that in general reasoning is incomplete if individuals are used in concept terms The following query is given w r t the above mentioned knowledge base given in the OWL file ex2 owl in the examples folder concept subsumes at most 1 file C Ralf Ind Examples ex1 owl R file C Ralf Ind Examples ex1 owl c NIL If dealing with nominals were no approximation i e if reasoning were complete then RacerPro would be able to prove a subsumption relationship because all r one of j implies at most 1 r RacerPro can download imported ontology documents Use the command owl read file lt filename gt to read an OWL file or use owl read document lt url gt to read an OWL resource given a URL You can manage multiple knowledge bases with RacerPro and you can load multiple on tologies into a single knowledge base For instance try Racer Systems GmbH amp Co KG http www racer systems c
230. nning Racer Systems GmbH amp Co KG http www racer systems com 210 active rules gt rule 1 get next set of rule consequences rule 1 gt instance alice mother related alice child of alice has child get current set of rule consequences rule 1 gt instance alice mother related alice child of alice has child choose current set of rule consequences rule 1 gt instance alice mother related alice child of alice has child get next set of rule consequences rule 1 gt instance betty mother related betty child of betty has child get next set of rule consequences rule 1 gt instance eve mother related eve child of eve has child get next set of rule consequences rule 1 gt instance doris mother related doris child of doris has child get next set of rule consequences rule 1 exhausted active rules gt NIL processed rules gt rule 1 all role assertions gt alice child of alice has child eve doris has sister doris eve has sister charles betty has sibling betty eve has child betty doris has child alice charles has child alice betty has child CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL applicable rules gt rule 1 execute or reexecute all queries rule 1 Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROC
231. nse terms for permitted usage We assume that RacerPro is on the search path of your operating system Racer Systems GmbH amp Co KG http www racer systems com 10 CHAPTER 2 USING RACERPRO 333 Racer and RacerPro are trademarks of Racer Systems GmbH amp Co KG 333 For more information see http www racer systems com 333 RacerPro comes with ABSOLUTELY NO WARRANTY use at your own risk 333 RacerPro is based on International Allegro CL Enterprise Edition 7 0 Oct 19 2004 13 28 333 Copyright C 1985 2004 Franz Inc Oakland CA USA All Rights Reserved 333 The XML RDF RDFS OWL parser is implemented with Wilbur developed 333 by Ora Lassila For more information on Wilbur see 333 http wilbur rdf sourceforge net 333 Found license file 333 Users rm ralf mac os x lic 333 This copy of RacerPro is licensed to 333 Ralf Moeller Hamburg University of Technology TUHH 333 Harburger Schlossstr 20 333 STS Group 33 21079 Hamburg 333 Deutschland 333 Initial license generated on 06 29 2005 12 35 for 1 8 1 333 Site Commercial on Mac OS X 333 This license is valid up to version 9 9 99 333 This license is valid forever 333 This is RacerPro for Ralf Moeller HTTP service enabled for http localhost 8080 TCP service enabled for http localhost 8088 If you do not have a valid license you are allowed to use RacerPro but some restrictions apply see the RacerPro web site f
232. nsequences must be chosen e In the set at a time mode all computed produced rule consequences can be added automatically to an ABox if nRQL is in add rule consequences automatically mode All computed rule consequences will be added to the ABox after the rule has terminated In case dont add rule consequences automatically mode is used the computed set of rule consequences will be memorized and later add chosen sets of rule consequences can be called on this rule to add the memorized rule consequences although nothing was really chosen here please read further e In the incremental tuple at a time mode the nRQL API provides appropriate func tions which can be used to implement you own rule application strategy As in the incremental query answering modi instead of returning the whole set of sets of rule consequences at once in a bunch nRQL will incrementally construct and return one set of rule consequences after the other one by one The next set of rule conse quences must be explicitly requested using get next set of rule consequences Then nRQL can be advised to memorize this computed set of rule consequences with choose current set of rule consequences In case nRQL is in add rule consequences automatically mode the chosen sets of rule consequences are added automatically to the ABox after the last set had been delivered Otherwise for example in case the rule was aborted the API function add chosen sets of r
233. nstance of owl Class asserted Hierarchy 0 1S A A Inferredtierarchy amp E Name Sameas Differentrrom annotations BeBe gt ow Thing cm CHAD owl Thing o Property value Lang family FEMALE family FEMALE family HUMAN v family HUMAN rdfs comment family MALE v family PERSON Y family MAN v familyMAN family BROTHER v O family BROTHER family FATHER famil UNCLE family UNCLE jA family FATHER Ser Y famiy PERSON E Y family PARENT Asserted Inferred mPropert at of mm Y family PARENT pali Asserted Conditions 6 6 a amp family FATHER d a flv NECESSARY amp SUFFICIENT v family MOTHER fami GRANDMOTHER i famiy GRANDMOTHER Y famiy WOMAN _ Y family WOMAN y family MOTHER family AUNT family GRANDMOTHER v family MOTHER v family SISTER family GRANDMOTHER family AUNT ZA gt family MALE wns oe ae ci a a hm Logic View Properties View Class Changed superclasses family AUNT Moved from family WOMAN to family SISTER family MAN Moved from owl Thing to farnily PERSON farnily PERSON Moved from owl Thing to family HUMAN family UNCLE Moved from family MAN to family BROTHER family WOMAN Moved from owl Thing to family PERSON C 9 Classification Results Figure 2 17 Prot g visualizes implicit subsumption relationships computed by RacerPro In Figure 2 17 it is indicated that RacerPro r
234. ntology1159352693 owl xmlns rdf http www w3 0rg 1999 02 22 rdf syntax ns xmlns xsd http www w3 org 2001 XMLSchema xmlns rdfs http www w3 0rg 2000 01 rdf schema xmlns owl http www w3 org 2002 07 owl xml base http www owl ontologies com Ontology1159352693 0w1 gt lt owl Ontology rdf about gt lt rdfs comment rdf datatype http www w3 org 2001 XMLSchema string gt Demo Ontology lt rdfs comment gt lt owl Ontology gt lt owl Class rdf ID class gt lt rdfs comment rdf datatype http www w3 0rg 2001 XMLSchema string gt This is a test class lt rdfs comment gt lt owl Class gt lt owl ObjectProperty rdf ID object property gt lt rdfs domain rdf resource class gt lt owl 0bjectProperty gt lt owl DatatypeProperty rdf ID datatype property gt lt rdfs range rdf resource http www w3 org 2001 XMLSchema int gt Racer Systems GmbH amp Co KG http www racer systems com 116 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL lt rdfs domain rdf resource class gt lt owl DatatypeProperty gt lt owl ObjectProperty rdf ID annotation object property gt lt rdf type rdf resource http www w3 org 2002 07 owl AnnotationProperty gt lt rdfs domain rdf resource class gt lt owl 0bjectProperty gt lt owl DatatypeProperty rdf ID annotation datatype property gt lt rdfs domain rdf resource class gt lt rdfs range rdf resource http www w3 org
235. number restrictions role hierarchies inverse roles and transitive roles In addition to these basic features RacerPro also provides facilities for algebraic reasoning including concrete domains for dealing with e min max restrictions over the integers e linear polynomial in equations over the reals or cardinals with order relations e equalities and inequalities of strings For these domains no feature chains can be supported due to decidability issues RacerPro supports the specification of general terminological axioms A T box may con tain general concept inclusions GCIs which state the subsumption relation between two concept terms Multiple definitions or even cyclic definitions of concepts can be handled by RacerPro RacerPro implements the HTTP based quasi standard DIG for interconnecting DL systems with interfaces and applications using an XML based protocol 4 RacerPro also implements most of the functions specified in the older Knowledge Representation System Specification KRSS for details see 17 Given a T box various kinds of queries can be answered Based on the logical semantics of the representation language different kinds of queries are defined as inference problems Racer Systems GmbH amp Co KG http www racer systems com 1 1 VIEWS ON RACERPRO 3 hence answering a query is called providing inference service As a summary we list only the most important ones here e Concept consis
236. ole fillers NIL told attribute fillers NIL told datatype fillers NIL annotation datatype property fillers NIL Vv Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 121 annotation property fillers NIL direct types to be computed describe query last gt query 1 accurate processed retrieve x lambda x describe ind x current abox x x top abox default retrieve x types x x top gt x i lambda x instantiators x current abox x c d top top retrieve x all types x x top gt x i lambda x instantiators x current abox x c d top top retrieve x all types flat x x top gt x i lambda x sort symbol name lessp flatten instantiators x current abox x c d top top Please note that the retrieve macro can be advised to no include the lambda expression in the query answer retrieve x all types flat x x top dont show lambdas p t gt x i c d top top It should be noted that the lambda expression can also be used in rule antecedences see below 6 1 3 3 MiniLisp Examples We present four examples which make MiniLisp more concrete and demonstrate its useful ness Racer Systems GmbH amp Co KG http www racer systems com 122 CHAPTER 6 THE NEW RACERPRO QUERY
237. ologies com Ontology1162148702 owl france y http www owl ontologies com Ontology1162148702 0wl germany P x http www owl ontologies com Ontology1162148702 0wl germany y http www owl ontologies com Ontology1162148702 0wl france begin verbatim It is interesting to oberserve that the RCC substrate cannot prove that tt paris and france are in tt dc relation actually they could be tt ec as well A coarser RCC relation specified with synonyms will also return Racer Systems GmbH amp Co KG http www racer systems com 192 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL tt paris and tt france try begin verbatim retrieve x Pry and x y adjacent disjoint Moreover note that the full annotations added by the mirror data substrate for the OWL objects is available for retrieval as well get data node description http www owl ontologies com Ontology1162148702 owl paris gt node name http www owl ontologies com Ontology1162148702 owl paris node label top http www owl ontologies com Ontology1162148702 owl city abox object abox individual owl object owl individual node successors NIL node predecessors http www owl ontologies com Ontology1162148702 0wl hamburg http www owl ontologies com Ontology1162148702 owl germany http www owl ontologies com Ontology1162148702 owl europe http www
238. om 6 1 THE NRQL LANGUAGE 179 0 only the told syntactic information from the ABox is exploited Thus only the instance assertions are analyzed and added to the label In nRQL mode 1 also taxonomic information can be added to the labels full reset gt okay full reset N enable data substrate mirroring v okay data substrate mirroring enabled racer read file nrql user guide examples family no signature racer OKAY v N enable smart abox mirroring v okay smart abox mirroring enabled retrieve nil x top gt t N get data node description alice node name alice node label top woman parent human mother person abox object abox individual node successors charles betty 80 node predecessors charles betty Vv Much more information about alice is available now Of course in a query one can always refer to the associated ABox individual of such a node so full description logic reasoning is always available Regarding the data edges created due to role membership assertions also the role hierarchy is taken into account get data edge description doris eve gt from node doris to node eve edge label has sister abox relationship Racer Systems GmbH amp Co KG http www racer systems com 180 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL abox role relationship has sibling Note that th
239. om 6 1 THE NRQL LANGUAGE 187 refined by considering whether the contained object is a a tangential proper part is tangentially contained i e touches the border from the inside or not In the former case TPP is used and in the latter case the non tangential proper part relationship NTPP The inverse relations are called TPPI and NTPPI Moreover disjunctions of these relations can be used to represent coarser or indefinite knowledge regarding the spatial relationship In general the RCC edges labels have the same CNF structure as ordinary edge labels However certain symbols denote RCC base relations The following edge labels will be recognized as RCC relations e either a single symbol in lowercase starting with a colon e g dr ppi or e a list of lists of such RCC symbols denoting a conjunction of disjunctions of RCC relations e g ec dc Note that dr ec dc pp ntpp tpp and ppi ntppi tppi Please also note that an edge label such as ntpp tpp would actually denote a conjunc tion and not a disjunction Since the RCC base relations have the so called JEPD jointly exhaustive pairwise disjoint property such a conjunction denotes the empty relation The RCC substrate supports is aware of the spatial characteristics inherent properties of the RCC relationships For example suppose you want to represent the spatial situation in which object a contains object b
240. om 76 CHAPTER 4 DESCRIPTION LOGIC MODELING WITH RACERPRO owl read document http www w3 org TR 2004 REC owl guide 20040210 wine rdf kb name dinner owl read document http www co ode org ontologies pizza 2005 05 16 pizza owl kb name dinner init nil for a delicious dinner Imported ontologies are automatically loaded from the corresponding web server Make sure you are connected to the Internet or use the mirror functionality if you are offline mirror lt url speci gt lt another url or local filename gt Examples mirror http www w3 org TR 2004 REC owl guide 20040210 wine rdf home users rm wine rdf mirror http www w3 org TR 2004 REC owl guide 20040210 wine rdf http localhost 8081 examples wine rdf Mirror statements might be placed in a RacerPro init file see the previous subsection In RacerPorter you can abbreviate the display of OWL names with the check box Simplify Names There is a known limitation The current version of the RDF reader of RacerPro ignores nodes with parseType Resource or parseType Literal This will be fixed in the next version In LRacer or RacerMaster you have to quote the symbol dinner Racer Systems GmbH amp Co KG http www racer systems com Chapter 5 Knowledge Base Management 5 1 Configuring Optimization Strategies The standard configuration of RacerPro ensures that only those computations are per formed that are required
241. ompute all implicit class subsumption relationships Reasoning services are provided when Prot g is connected to RacerPro This can be easily accomplished The role hierarchy of the Family example is shown Figure 2 15 Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO 35 ee8 family 1 Prot g 3 1 beta file Users rm Desktop family 1 pprj OWL Files owl or rdf File Edit Project OWL Code Window Tools Help DCA S as FER OS D EE a gt ZK prot g OWLClasses mm Properties Forms Individuals Metadata PROPERTY BROWSER PROPERTY EDITOR Po T For Project family 1 For Property W family HAS DESCENDANT instance of owl ObjectProperty owl TransitiveProperty E Properties of af t cf cf my Name Equivalents Sameas DifferentFrom Annagioni EC N v mm family HAS DESCENDANT family HAS DESCENDANT Q Property value Lang m family HAS CHILD m family HAS GENDER rdfs comment P gt mm familyHAS SIBLING Domain amp e Range G amp amp owl Thing C Functional InverseFunctional C Symmetric v Transitive Inverse m p a Super properties mu Sb b soe Figure 2 15 Display of the role hierarchy with Prot g 0909 family 1 Prot g 3 1 beta file Users rm Desktop family 1 p
242. on are used For example retrieve x y name of y age of y tv 1 tv 2 and x abox individual x y has child y woman y age of y abox attribute relationship age y name of y abox attribute relationship name age of y abox concrete domain object name of y abox concrete domain object age of y tv 1 abox told value relationship name of y tv 2 abox told value relationship tv 1 abox concrete domain value predicate lt 90 P tv 2 abox concrete domain value predicate search tty gt x alice y betty name of y betty name age of y betty age tv 1 50 tv 2 Betty Moreover the amount of information which is mirrored can be controlled Let us consider the node alice again get data node description alice gt node name alice node label top mother abox object abox individual node successors charles betty 80 alice name alice age node predecessors charles betty So the node label contains the told concept membership information top mother This information is computed and added to the labels according to the current nRQL mode see the documentation of set nrql mode in the Reference Manual The nRQL mode controls the amount of completeness of the representation For example using nRQL mode Racer Systems GmbH amp Co KG http www racer systems c
243. onal set whereas in the pervious query neg was applied to and returned a two dimensional set This is easy to verify using describe query again retrieve x alice x has child gt x charles x betty N describe query last query 7 accurate processed retrieve x and same as alice ano1 alice alice ano1 x has child abox v smith family retrieve x neg alice x has child Racer Systems GmbH amp Co KG http www racer systems com 142 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL gt x eve x doris x charles x betty x alice describe query last gt query 8 accurate processed retrieve x or and not same as alice ano1 alice top x and top alice not alice anol x has child abox smith family Moreover regarding individuals in query heads we have the same situation as already dis cussed for the unary atoms they are replaced with representative variables and same as conjuncts This is the standard transformation Negated individuals thus behave like vari ables retrieve alice alice x has child gt alice alice retrieve alice meg alice x has child gt alice eve alice doris alice charles alice betty alice alice Again if this behavior is not desired simply avoid using individuals in NAF negated atoms and add same as and or n
244. onomy node gt x top x woman with children x woman x bottom Note that nRQL has selected woman with children and not mother as the representative concept However representative concepts are only needed in order to reduce the number of query answer tbox retrieve x y x y has child gt x woman y mother x woman y woman with children x mother y bottom x mother y bottom x woman with children y bottom x woman with children y bottom x top y woman x top y woman tbox retrieve x y and x y has child x taxonomy node y taxonomy node gt x top y woman x woman with children y bottom x woman y woman with children It is easy to retrieve the synonyms of concept tbox retrieve y mother y has synonym gt y mother y woman with children tbox retrieve y and mother y has synonym neg same as mother y gt y woman with children We can also ask for the concepts which have no synonyms Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 163 tbox retrieve x neg project to x and x y has synonym neg same as x y gt x woman Moreover the two head projection operator concept synonyms and describe are under stood tbox retrieve x concept synony
245. onstructs for retracting T box axioms see the function forget statement However complete reclassification may be necessary in order to answer queries Retracting axioms is mainly useful if the RacerPro server is used With retracting there is no need to delete and retransfer a knowledge base T box RacerPro also offers constructs for retracting A box assertions see forget forget concept assertion forget role assertion and friends If a query has been answered and some assertions are retracted then RacerPro might be forced to compute the index structures for the A box again realization i e after retractions some queries might take some time to answer Note that may queries are answered without index structures at all see also Section 5 1 RacerPro also supports incremental additions to A boxes i e assertions can be added even after queries have been answered However the internal data structures used for answering queries are recomputed from scratch This might take some time If an A box is used for hypothesis generation e g for testing whether the assertion 7 C can be added without causing an inconsistency we recommend using the instance checking inference service If individual instance i not C returns t i C cannot be added to the A box Now let us assume we can add i C and afterwards want to test whether i D can be added without causing an inconsistency In this case it might be faster not to add i C directly b
246. ontology Certainly hidden individuals can be marked as such with special concept names and in queries they might explicitly be excluded by conjoining the negation of the marker concept automatically to the query concept However from an implementation point of view this can be provided much more efficiently if the mechanism is built into the retrieval machinery of RacerPro Racer Systems GmbH amp Co KG http www racer systems com 80 CHAPTER 5 KNOWLEDGE BASE MANAGEMENT define concrete domain attribute isbn define concrete domain attribute number of copies sold implies implies implies implies implies implies book document article document computer science document document computer science book and book computer science document compiler construction book computer science book and min number of copies sold 3000 computer science document computer science best seller In order to manage assertions about specific documents an A box current documents is defined with the following statements The A box current documents is the current A box to which subsequent statements and queries refer The set of subscriptions w r t the current A box is initialized i e wIth init subscription you can instruct the server to delete all previous subscriptions in abox current documents document ontology init subscriptions With the following set of statements five document ind
247. or details If you have a valid license your own name will be mentioned in the output of course The following forms are found in the file family racer in the examples folder 333 initialize the T box family in tbox family 333 Supply the signature for this T box Racer Systems GmbH amp Co KG http www racer systems com 2 1 SAMPLE SESSION WITH RACERPORTER 11 signature atomic concepts person human female male woman man parent mother father grandmother aunt uncle sister brother roles has child parent has descendant has descendant transitive t has sibling has sister parent has sibling has brother parent has sibling has gender feature t individuals alice betty charles doris eve Domain amp range restrictions for roles implies top all has child person implies some has child top parent implies some has sibling top or sister brother implies top all has sibling or sister brother implies top all has sister some has gender female implies top all has brother some has gender male Axioms for relating concept names implies person and human some has gender or female male disjoint female male implies woman and person some has gender female implies man and person some has gender male equivalent parent and person some has child person equivalent mother and woman parent equivalent father and man
248. owing options are important to control the behavior or the console window e The option p make the console windows persistent i e you have to explicitly close the window This is useful to read error messages e The option c instructs RacerPro not to open the console window This option is useful for file based operation of RacerPro batch mode 2 3 How to Send Bug Reports Although RacerPro has been used in some application projects and version 1 9 has been extensively tested it might be the case that you detect a bug with respect to a particular knowledge base In this case please send us the knowledge base together with the query as well as a description of the RacerPro version and operating system It would be helpful if the knowledge base were stripped down to the essential parts to reproduce that bug Before submitting a bug report please make sure to download the latest version of RacerPro Racer Systems GmbH amp Co KG http www racer systems com 2 4 RACERPORTER 25 Sometimes it might happen that answering times for queries do not correspond adequately to the problem that is to be solved by RacerPro If you expect faster behavior please do not hesitate to send us the application knowledge base and the query or queries that cause problems In any case get the latest version of RacerPro first As a registered RacerPro user you may send your questions and bug reports to the following e mail address sup
249. owl ex1 owl defin ing one instance michael of class person and one instance book123 of class book lt xml version 1 0 gt lt rdf RDF xmlns rdf http www w3 0rg 1999 02 22 rdf syntax ns xmlns xsd http www w3 org 2001 XMLSchema xmlns rdfs http www w3 0rg 2000 01 rdf schema xmlns owl http www w3 org 2002 07 owl xmlns http www owl ontologies com unnamed owl xml base http www owl ontologies com unnamed owl gt lt owl Ontology rdf about gt lt owl Class rdf ID person gt lt owl Class rdf ID bo00k gt lt person rdf ID michael gt lt book rdf ID b00k123 gt lt rdf RDF gt With nRQL we can easily retrieve the instances of the OWL classes http www owl ontologies com unnamed owl person and http www owl ontologies com unnamed owl book full reset gt okay full reset owl read file Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 97 nrql user guide examples owl ex1 0w1 gt home mi wessel nrql user guide examples owl ex1 owl get namespace prefix gt http www owl ontologies com unnamed owl retrieve x x http www owl ontologies com unnamed owl person gt x http www owl ontologies com unnamed owl michael retrieve x x http www owl ontologies com unnamed owl book gt x http www owl ontologies com unnamed owl boo0k123 retrieve x x book gt
250. ox individual e for each concrete domain object and e for each concrete domain value Edges are created in the mirror for Racer Systems GmbH amp Co KG http www racer systems com 176 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL e related as well as for e constrained axioms Let us inspect some nodes get data node description 80 gt node name 80 node label abox object abox concrete domain value 80 node successors NIL node predecessors alice alice age get data node description alice gt node name alice node label top mother abox object abox individual node successors charles betty 80 alice name alice age node predecessors charles betty get data node description alice name gt node name alice name node label abox object abox concrete domain object node successors Alice node predecessors alice get data node description Alice gt node name Alice node label abox object abox concrete domain value Alice node successors NIL node predecessors alice name We see that there are various kinds of nodes nodes for ABox individuals alice nodes for concrete domain objects alice name and nodes for told concrete domain told datatype values 80 Alice Appropriate markers are added to the node labels in order to make them self describing see node label entries in
251. p Metadata 33 SWRL Rules Name Expression Rule 1 2 French x Wine drinker x Rule 2 2 Italian x Wine drinker x Figure 2 37 Rules for French and Italian If the rules in Figure 2 37 are specified the ontology would be inconsistent with first order semantics There are two worlds to consider Assume john is a French citizen then he must be an instance of Wine drinker due to Rule 1 see Figure 2 37 But he is declared to be an instance of Not wine drinker In the other possible world john is an instance of Italian Again due to the rules he must be a Wine drinker which results in a contradiction In the rule semantics one can neither prove that john is an instance of Italian nor can one prove that john is an instance of French Thus the rules are not applied and the ontology remains consistent RacerPro applies the rule semantics in version 1 9 This can be seen as an advantage or as a disadvantage In order to achieve the same effect as in the first order semantics in this case a rule with precondition Not wine drinker could be added Note also that simple rules such as those shown in Figure 2 37 can be represented as concept axioms in Prot g itself Racer Systems GmbH amp Co KG http www racer systems com 54 CHAPTER 2 USING RACERPRO Racer Systems GmbH amp Co KG http www racer systems com Chapter 3 RacerPro Knowledge Bases In description logic systems a
252. parent equivalent grandmother and mother some has child some has child person equivalent aunt and woman some has sibling parent equivalent uncle and man some has sibling parent equivalent brother and man some has sibling person equivalent sister and woman some has sibling person instance alice mother related alice betty has child related alice charles has child instance betty mother related betty doris has child related betty eve has child instance charles brother related charles betty has sibling instance charles at most 1 has sibling related doris eve has sister Racer Systems GmbH amp Co KG http www racer systems com 12 CHAPTER 2 USING RACERPRO related eve doris has sister Start RacerPorter by double clicking the program icon or type RacerPorter as a command in your shell Press the button Connect to connect the graphical interface to the RacerPro reasoning engine Make sure RacerPro is already started Then load a file into RacerPro by pressing the button Load or by selecting Load in the menu File on Mac OS X Ga eee RacerPorter Shell Cons TBoxes ABoxes Taxo Role Netw Conc Roles Indivi Asser Queries Rules TBox ABox Concept Role Individual Query Rule lify Na Command Connect Profiles s Clear Log F eset New Open RacerPro Log lij asuueseesuquuuuesquutiseze
253. pile this file Test java with the package jracer see the JRacer directory in your RacerPro distribution in the same directory or on the Java CLASS PATH LRacer LRacer is the API for Common Lisp to access all services of RacerPro in a convenient way LRacer is provided with source code You can download the latest version from http www racer systems com products download LRacer provides all functions and macros described in the RacerPro user s guide and reference manual directly from Common Lisp Thus from Common Lisp you do not send strings to the server directly but use stub functions or macros which internally communicate with the RacerPro server interpret the results and provide them as Common Lisp data structures In this sense LRacer is not more powerful than JRacer but a little bit more convenient Note the difference between LRacer and RacerMaster Although LRacer provides the same functionality all functions calls are sent to a RacerPro server via the TCP socket interface whereas for RacerMaster there is no such overhead The advantage of LRacer is that you can run RacerPro on a powerful Racer Systems GmbH amp Co KG http www racer systems com 22 CHAPTER 2 USING RACERPRO server machine and develop your application on a less expensive portable computer Let us assume the LRacer directory is stored under Lracer Start your favorite Com mon Lisp system and evaluate load Lracer lracer sysdcl lisp
254. port racer systems com If you want to submit a bug report or a question about a certain behavior of RacerPro please attach the logfile to your e mail Of course the logfile should cover the session in which the error occurred Logging including a stack backtrace in case of an error is enabled by starting the executable as follows under Windows RacerPro exe log lt filename gt debug or under Unix RacerPro log lt filename gt debug Please include at least this information in your correspondence e Your personal name and the name of your organization e Your operating system e Your contact data including telephone number The logfile see above e The RacerPro build number e Your transaction ID or the short license string A description of your problem and the environment where it occurred We may need additional information about your setup or some of the data files you process to simulate the reported error condition You will greatly decrease our response time if you help us by providing such information on request Of course we will try to work on your issue as soon as possible However due to the probably existing time lag between your location and ours or due to the existing work load our response may take up to two business days We will acknowledge the receipt of your inquiry and in most cases give you further instructions or an estimate of the processing time 2 4 RacerPorter With RacerPorter you can intera
255. prj File Edit Project OWL Code Window Tools Help Heh BE wa Y Aol Db GE db OWLClasses Im Properties Forms Individuals Metadata SUBCLASS RELATIONSHIP lt gt mi CLASS EDITOR For Project family 1 For Class famil GRANDMOTHER instance of owl Class Figure 2 16 Prot g icon panel Y The Prot g icon panel provides for buttons labeled with C I They are used to check the consistency coherence of the ontology classify the ontology and to compute the inferred types of individuals respectively Press the C button in the tools bar to let RacerPro check for unsatisfiable concept names in the current knowledge base and find implicit subsumption relationships between concept names Racer Systems GmbH amp Co KG http www racer systems com 36 CHAPTER 2 USING RACERPRO le ee family 1 Prot g 3 1 beta file Users rm Desktop family 1 pprj OWL Files owl or rdf File Edit Project OWL Code Window Tools Help Dee Bf ug VY AOR 265 E EE db lt lt prot g ow classes mu Properties Forms Individuals Metadata SUBCLASS RELATIONSHIP SUBCLASS RELATIONSHIP O x CLASS EDITOR EE For Project family 1 For Project family 1 For Class owl Thing i
256. pt their code to use the HTTP based DIG interface 2 2 5 Options for the RacerPro Server Various options allow you to control the behavior of RacerPro Under Windows we have to distinguish between RacerPro options and window manager options see below Under Windows you have to use the separator even if there are no window manager options used e Use the option h to get the list of possible options and a short explanation Use this option in combination with the window manager option p see below e As indicated above the options f lt filename gt can be used for reading knowledge bases from a file The extension of jfilenamej is used to discriminate the syntax to be used possible extensions for corresponding syntaxes are racer owl rdfs rdfs In all other cases Racer syntax is expected If the extension of your input file does not have an appropriate extension you can specify the syntax with the options owl dig rdfs instead of f e Use the option q lt filename gt to specify a file with queries extension racer only e p lt port number gt specifies the port for TCP connections e g for LRacer and JRacer see above e http lt port number gt specifies the port for HTTP connections e g for the DIG interface see above e httplogdir lt directory gt specifies the logging directory see above e nohttpconsolelog disables console logging for HTTP connections see above e Processing knowledge
257. ptional mode 1 check abox consistency gt mode 2 run lubm tests original lubm n check abox consistency check abox consistency mode mode Racer Systems GmbH amp Co KG http www racer systems com 258 APPENDIX D LUBM BENCHMARK Racer Systems GmbH amp Co KG http www racer systems com Bibliography 1 2 F Baader D Calvanese D MacGuinness D Nardi and P F Patel Schneider editors The Description Logic Handbook Cambridge University Press Cambridge UK 2003 F Baader I Horrocks and U Sattler Description logics as ontology languages for the semantic web In Festschrift in honor of J rg Siekmann Lecture Notes in Artificial Intelligence Springer 2003 2003 Franz Baader Bernhard Hollunder Bernhard Nebel Hans J rgen Profitlich and En rico Franconi An empirical analysis of optimization techniques for terminological representation systems or making KRIS get a move on In B Nebel W Swartout and C Rich editors Principles of Knowledge Representation and Reasoning Proceed ings of the 3rd International Conference pages 270 281 San Mateo 1992 Morgan Kaufmann S Bechhofer P Crowther and R M ller The description logic interface In D Cal vanese G De Giacomo and E Franconi editors International Workshop on Descrip tion Logics pages 196 203 September 2003 M Buchheit F M Donini and A Schaerf Decidable reasoning in terminological knowledge
258. r This atom simply returns the complement set of x grandmother For any concept C the body union x C neg x C is equivalent to x top N retrieve x union x grandmother neg x grandmother x alice x eve x doris x charles x betty v N retrieve x x top x alice x betty x charles x doris x eve Vv Note that classical negation is of course also available since RacerPro offers concept negation in its concept language retrieve x x not grandmother gt x charles RacerPro is able to prove that charles is an instance of not grandmother since carles is aman and grandmothers are woman Moreover man and woman are disjoint Note that for any concept C the atom x not C always returns a subset of neg x C Thus as already mentioned the following query is not equivalent to x top retrieve x union x grandmother x not grandmother gt x alice x charles It even possible to combine classical with NAF negation in a query retrieve x meg x not grandmother gt x eve x doris x betty x alice So we have asked for those individuals for which RacerPro cannot prove that they are instances of not grandmother Note that this is again the complement of retrieve x x not grandmother gt x charles 6 1 4 3 2 Negation as Failure for U
259. r by definition although the actual number of sold copies is not known to RacerPro describe individual document 8 DOCUMENT 8 ASSERTIONS DOCUMENT 8 COMPUTER SCIENCE BEST SELLER ROLE FILLERS NIL TOLD ATTRIBUTE FILLERS ISBN 8234567 DIRECT TYPES COMPUTER SCIENCE BEST SELLER Racer Systems GmbH amp Co KG http www racer systems com 84 CHAPTER 5 KNOWLEDGE BASE MANAGEMENT The following subscription queries indicate that the query concept must not necessarily be a concept name but can be a concept term state subscribe client 4 equal isbn 7234567 RacerPro returns the following information CLIENT 4 DOCUMENT 7 Notice again that subscriptions might be considered when new information is added to the A box state subscribe client 5 equal isbn 10234567 The latter statement returns NIL However the subscription is considered if at some time point later on a document with the corresponding ISBN number is introduced and pub lished state instance document 10 document constrained document 10 isbn 10 isbn constraints equal isbn 10 10234567 publish document 10 CLIENT 5 DOCUMENT 10 This concludes the examples for the publish subscribe facility offered by the RacerPro sys tem The publish subscribe mechanism provided with the current implementation is just a first step This facility will be extended significantly Future versions will include optimiza tion te
260. r of requested tuples was reached 6 2 2 The Life Cycle of a Query A query has a life cycle It is created and prepared parsed and compiled then made active computes tuples eventually goes to sleep will be awoken again compute some more tuples and eventually becomes inactive its query answering thread dies see Figure 6 2 Racer Systems GmbH amp Co KG http www racer systems com 206 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL Active Queries Terminated Queries Prepared Queries retrieve prepare abox query execute query abort query reexcute query reprepare query Figure 6 2 Lifecycle of a Query Even after its tread is terminated the query is still present as a query object and can be reexecuted RacerPorter provides the Queries tab which can be used to inspect and manage the queries as well as their current states Moreover the function describe all queries and describe query provide status information about queries We can distinguish the following lists sets of queries according to their status 1 The list of all queries The corresponding API function all queries returns this list As long as a query appears on this list its ID is recognized by the API functions Use delete query delete all queries etc to remove queries from this list 2 Queries which are ready to run but have not been started yet The corresponding API funct
261. r of the has age attribute is given as a told value in the KB it can be returned by the query together with the individual Racer Systems GmbH amp Co KG http www racer systems com 89 nRQL also supports binary concrete domain predicates not that gt 18 is a unary concrete domain predicate For example we can retrieve all pairs of persons having the same age with the constraint query atom x y constraint age age Please note that the available retrieval predicates are from the concrete domains offered by RacerPro Thus for the sake of decidability in the RacerPro concept lan guage their number is rather limited e g there is no concrete domain predicate which checks whether a string contains a certain fixed substring However such additional predicates are often needed for query formulation A pragmatic solution to this missing predicates problem is to enable the so called mirror data substrate and use the additional retrieval facilities offered by the data substrate query atoms see below Here you will find many useful predicates But unlike the concrete domain predicates these data substrate predicates only work on told data values Moreover a third pragmatic solution to the missing predicates problem is to use the MiniLisp expression language see below to specify arbitrary custom retrieval filter predicates nRQL offers a projection operator project to for query bodies In combinat
262. raint age age gt age 1 age 2 40 gt x alice y doris x alice y eve The complex predicate gt age 1 age 2 40 is best understood as a lambda expression with two formal parameters age 1 and age 2 which are bound to the actual arguments supplied by the age of x and the age of of y Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 105 A age1 agez2 e age gt agez 40 age x age y Thus A age1 age2 e ager gt agez 40 is applied to the actual arguments age x and age y and either returns TRUE or FALSE Note that the formal parameters here age age2 are computed from the attribute names The suffixes 1 and 2 are added just in case the same attribute is used in order to be able to differentiate the two in the body of the predicate 6 1 1 3 3 Querying OWL KBs with Constraint Query Atoms In the OWL realm the equivalent of concrete domain attributes are called OWL datatype properties Suppose we add to owl ex2 owl two datatype property declarations for the properties age and name and specify appropriate fillers for betty and alice as follows see file owl ex3 owl lt owl DatatypeProperty rdf ID age gt lt rdfs domain rdf resource person gt lt rdf type rdf resource http www w3 org 2002 07 owl FunctionalProperty gt lt rdfs range rdf resource http www w3 org 2001 XMLSchema int gt lt owl DatatypeProperty gt
263. rate edge 149 data substrate edge query expression 152 data substrate label 149 data substrate node 149 data substrate node query expression 152 data substrate predicate 153 data substrate query expression 152 deadlock prevention 185 defined query 133 defined query and NAF 135 defined query and PROJECT TO 135 DIG 22 disjoint concepts 59 disjunctive query 120 domain restriction 61 eager mode 174 editions 5 equal role 99 exists restriction 57 expensive query 182 expensive rule 182 expensive tuple 182 explicit role filler query 101 extended OWL query 154 158 feature 60 feature chains in queries 104 GCI 59 graphical client interfaces 29 head projection operators 110 hybrid ABox individual 150 Racer Systems GmbH amp Co KG http www racer systems com INDEX 263 hybrid OWL query 154 158 hybrid query 151 154 hybrid representation 149 implicit role filler query 101 incomplete mode 179 incremental mode 172 incremental query processing 173 individual 92 inference modes 66 injective variable 92 installation 6 JRacer 20 lazy mode 174 local closed world assumption 85 LRacer 21 macro query 133 marker 157 mirror data substrate 154 mirror of an ABox 154 mirror substrate 154 mirror substrate marker 157 mirror substrate query 154 mirroring an ABox 154 mirroring an OWL file 154 mirroring OWL documents 158 NAF 124 NA
264. rce RacerPro to apply the unique name assumption Sometimes for debugging purposes it is useful to inspect the commands your appli cation sends to RacerPro Specify log lt filename gt to print logging information into a file see also the command logging on to switch logging on dynamically Specify temp lt directory gt if you would like to change the default directory for temporary files the default is temp under Unix based systems and the value of the environment variable TEMP or TMP under Windows Sometimes using the option silent is useful if you want to suppress any diagnostic output RacerPro supports the DIG protocol with some extensions For instance RacerPro in terprets the DIG specification generated by Prot g in such a way that DIG attributes are treated as datatype properties in order to match the semantics of OWL Thus by default DIG attributes do not imply at most one filler If you have an application that relies on DIG 1 1 specify the option dig 1 1 to instruct RacerPro to obey the original semantics of DIG attributes In case you have problems with the license file license racerlicense you can start RacerPro with the option license lt license file gt This prints informa tion about the file lt license file gt More detailed information is printed with dump license info lt filename gt Note again that under Windows RacerPro options have to be separated with from window manager options The foll
265. representation systems Journal of Artificial Intelligence Research 1 109 138 1993 V Haarslev R M ller and M Wessel The description logic ALCNH p extended with concrete domains A practically motivated approach In R Gor A Leitsch and T Nipkow editors Automated Reasoning First International Joint Conference IJCAR 01 Siena Italy June 18 23 2001 Proceedings volume 2083 of Lecture Notes in Artificial Intelligence pages 29 44 Berlin 2001 Springer Verlag I Horrocks Optimising Tableaux Decision Procedures for Description Logics PhD thesis University of Manchester 1997 I Horrocks U Sattler and S Tobies A PSPACE algorithm for deciding ALCT g satisfiability LTCS Report 98 08 LuFg Theoretical Computer Science RWTH Aachen Germany 1998 I Horrocks U Sattler and S Tobies Reasoning with individuals for the description logic shiq In David MacAllester editor Proceedings of the 17th International Confer ence on Automated Deduction CADE 17 number 1831 in Lecture Notes in Computer Science pages 482 496 Germany 2000 Springer Verlag Racer Systems GmbH amp Co KG http www racer systems com 260 BIBLIOGRAPHY 10 I Horrocks U Sattler and S Tobies Reasoning with individuals for the descrip tion logic SHIQ In David MacAllester editor Proceedings of the 17th International Conference on Automated Deduction CADE 17 number 1831 in Lecture Notes in Computer Science German
266. retrieve a b and a woman a b has child abox smith family Please note that the formal objects arguments used in the query definition here x y are renamed to match the actual objects arguments here a b Obviously the number of actual parameters supplied must always match the number of formal parameters in the definition However if one is not interested in the bindings for a certain formal parameter then one can simple use NIL as an actual parameter retrieve a substitute mother of a nil gt a alice a betty N retrieve a a nil mother of a alice a betty v describe query last query 5 accurate processed retrieve a and a woman a y anol ano2 has child abox Vv smith family A defined query may be referred to at any position in a query body for example retrieve a neg a nil mother of gt a charles a eve a doris a betty a alice describe query last Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 149 gt query 8 accurate processed retrieve a or and not a woman top y anol ano2 not a y anol ano2 has child abox smith family However in most cases you probably want retrieve a neg project to a a nil mother of gt a eve a doris a charles instead see discu
267. ricciosa pt proper ie LH gt O hasBase single PizzaBase 4 v 0 hasTopping TomatoTopping HamTop PeperonataTopping AnchoviesTopping 8 CaperTopping s 7 Spie Tonino vr Doisjoints GQ a al me C FourSeasons n QuattroFormaggi D wv 7 AmericanHot Logic View Properties View multiple PizzaT T Added InterestingPizza CheeseyPizza SpicyPizzaEquivalent MeatyPizza SpicyPizza Added InterestingPizza CheeseyPizza SpicyPizzaEquivalent NonVegetarianPizza SpicyPizza InterestingPizza VegetarianPizzaEquivalent1 VegetarianPizzaEquivalent2 Figure 2 21 Some classes are inconsistent in pizza owl Racer Systems GmbH amp Co KG http www racer systems com CHAPTER 2 USING RACERPRO pizza owl Prot g 3 1 1 file Applications Protege_3 1 examples pizza pizza owl pprj OWL Files owl or rdf ajet mems lt lt prot g SUBCLASS RELATIO For Project pizza owl Asserted Hierarchy owl Thing Y DomainConcept Country IceCream v Pizza CheeseyPizza InterestingPizza MeatyPizza Y NamedPizza American AmericanHot Cajun Capricciosa Caprina Fiorentina FourSeasons FruttiDiMare Giardiniera LaReine For Project pizza owl owl Thing Do
268. ris objects temp eve temp doris constrained eve temp eve temperature fahrenheit constrained doris temp doris temperature celsius constraints temp eve 102 56 temp doris 39 5 For instance this states that eve is related via the attribute temperature fahrenheit to the object temp eve The initial constraint temp eve 102 56 specifies that the object temp eve is equal to 102 56 Now asking for the direct types of eve and doris reveals that both individuals are instances of human with fever In the following A box there is an inconsistency since the temperature of 102 56 Fahrenheit is identical with 39 5 Celsius constrained eve temp eve temperature fahrenheit constrained doris temp doris temperature celsius constraints temp eve 102 56 temp doris 39 5 gt temp eve temp doris Racer Systems GmbH amp Co KG http www racer systems com 64 CHAPTER 3 RACERPRO KNOWLEDGE BASES We present another example that might be important for many applications dealing with dates The following declarations can be processed with Racer The predicates divisible and not divisible are defined for natural numbers and are reduced to linear inequations internally define concrete domain attribute year type cardinal define concrete domain attribute days in month type cardinal implies Month and gt days in month 28 lt days in month 31 equivalent month inleapyear and Month divisi
269. rmat t done let t2b get internal real time when check abox consistency format t ABox consistency checking time abox consistent format t done let t2c get internal real time format t Compute index structures time prepare racer engine format t done let t2d get internal real time Allegro gc t format t Load 20T 4F Preparation 20T 4F Consistency 20T 4F Index 20T 4F7 tib tia internal time units per second t2b t2a internal time units per second if check abox consistency Racer Systems GmbH amp Co KG http www racer systems com 253 t2c t2b internal time units per second 0 t2d t2c internal time units per second defun run lubm benchmark benchmark function check abox consistency amp optional univs 0 min deps 0 max deps min deps loop for depsi from if null min deps 0 min deps to if or null max deps not zerop univs 0 max deps do format t Initializing 7 prepare lubm data n universities check abox consistency univs if or null max deps not zerop univs nil deps1 format t Querying 7 funcall benchmark function univs if or null max deps not zerop univs nil deps1 values defun original lubm universities departments 333 Query 1 benchmark universities departments 1 and x http
270. ro server is an executable file available for Linux Mac OS X Solaris 2 and Windows XP It can be started from a shell or by double clicking the corresponding pro gram icon in a graphics based environment For instance the Windows version is shown in Figure 2 8 RacerPro Welcome to RacerPro Version 1 9 6 2005 11 12 Racer Renamed Abox and Concept Expression Reasoner Supported description logic ALCQHIF D Supported ontology web language subset of OWL DL no so called nomi Copyright C 2604 2065 by Racer Systems GmbH amp Co KG All rights reserved See license terms for permitted usage Racer and RacerPro are trademarks of Racer Systems GmbH amp Co KG For more information see http wwu racer systems com RacerPro comes with ABSOLUTELY NO WARRANTY use at your own risk RacerPro is based on International Allegro CL Enterprise Edition 7 6 Oct 19 2004 13 28 Copyright C 1985 2004 Franz Inc Oakland CA USA All Rights R The XML RDF RDFS OWL parser is implemented with Wilbur developed by Ora Lassila For more information on Wilbur see http wilbur rdf sourceforge net Found license file c Ralf ralf x86 win32 lic This copy of RacerPro is licensed to Ralf Moeller Hamburg University of Technology TUHH Harburger Schlossstr 26 STS Group 21079 Hamburg Deutschland Initial license generated on 66 29 2065 12 35 for 1 8 1 Site Commercial on X86 Win32 This license is valid up to version 9 9 99
271. role is always the Current Role as shown in the Role text field in the state display The other list panes only allow a single selection In the Queries or Rules tab select a nRQL query from the list then use the buttons to apply a command on the selected query Note that the current query or rule is shown in the Query or Rule field in the state display at the top 2 4 4 Known Problems Note that the JPEG image shown in the About tab will only work on Linux if you have user lib libImlib so installed SuSE ln s opt gnome lib libImlib s0 1 9 14 user lib libImlib so 2 5 Other Graphical Client Interfaces for RacerPro In this section we present open source Java based graphical client interfaces for the RacerPro Server The examples require that RacerPro be started with the option u The first interface is the RICE system Afterwards a short presentation of the Prot g system with RacerPro as a backend reasoner is given Then the coordinated use of Prot g and RICE or RacerPorter is sketched 2 5 1 RICE RICE is an acronym for RACER Interactive Client Environment and has been developed by Ronald Cornet from the Academic Medical Center in Amsterdam RICE is provided with source code The executable version is provided as a jar file Newer versions of RICE can be found at http www b1g systems com ronald rice In order to briefly explain the use of RICE let us consider the family example again First start the RacerPro server As
272. rql equal role role query atoms manually e g we can ask if there is anyone who is not a child of alice retrieve alice and neg alice x has child same as alice alice gt alice alice However everyone is a descendant of alice retrieve alice x and neg alice x has descendant same as alice alice neg x alice nrql equal role gt NIL Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 143 6 1 4 4 A Note on Boolean Complex Queries So far we have introduced the and union and neg query constructors nRQL allows for the orthogonal composition of boolean query bodies For processing purposes nRQL transforms queries into Negation Normal form NNF Ina NNF query the neg operator appears only in front of query atoms Moreover for optimization purposes the queries are then brought into Disjunctive Normal Form DNF This transformation might result in an exponential blowup in query size We would like to inform the user of this potential performance pitfall The semantics of the original query is of courses preserved by the transformations 6 1 4 5 PROJECT TO The Projection Operator for Query Bodies The projection operator is needed in combination with neg We have already discussed some example in the previous sections which made use of the projection operator Why is the operator needed The NAF negation operator neg is d
273. ry mother of x y and x woman x y has child Vv N all queries gt NIL Note that a defined query is not created as a query object at definition time thus all queries returns NIL The list x y in the mother of definition is called the list of formal arguments of the definition it is a comparable to a query head however it differs from a query head because only objects variables and individuals are permitted no head projection operators The same restriction applies to the object list of the project to constructor The defined query named mother of can subsequently be used as follows N retrieve x y x y mother of x alice y betty x alice y charles x betty y doris x betty y eve Vv N describe query last gt query 2 accurate processed retrieve x y and x woman x y has child abox smith family Another way to reuse a defined query is by means of the substitute keyword Racer Systems GmbH amp Co KG http www racer systems com 148 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL A defined query can be reused resp referenced and its body inserted with the substitute keyword retrieve a b substitute mother of a b gt a alice b charles a alice b betty a betty b eve a betty b doris describe query last gt query 4 accurate processed
274. s expressions can be used in concept query atoms retrieve x x and woman some has child top gt x betty x alice Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 95 6 1 1 1 1 Terminology Query Head and Query Body As already mentioned the query retrieve x x woman has the head x and the body x woman The body of a query specifies the retrieval conditions and the head specifies the format of the query result answer tuples Each binding list lists a number of variable value pairs A binding list can also be seen as a component wise named tuple The head may also contain individuals However the set of objects mentioned in the query head must be a subset of the set of objects used in the query body An empty query head is permitted as well see below this will result in a boolean query which only returns TRUE or FALSE 6 1 1 1 2 Boolean Queries A boolean query only returns TRUE or FALSE Simply use an empty head to check whether there are any woman and aunts at all retrieve nil x woman gt t retrieve nil x aunt gt NIL Note that t means TRUE and NIL means FALSE 6 1 1 1 3 Queries with individuals Individuals can be used in queries as well retrieve nil betty woman gt t retrieve nil janice woman ERROR Undefined individual name janice in ABox smith family v retrieve nil betty aunt
275. s Each nRQL query body is associated with a so called object vector which is a sequence list of lexicographic ordered objects variables and individuals mentioned in the query body So the object vector of x woman is lt x gt and lt y gt for y woman Moreover lt x y gt for and x man y woman and and x woman y man In order to check whether the query a is more specific than query b the variables in b are renamed to match the names of the variables in a according to their positions in the object vectors So we are not considering all possible ways to rename the variables but only consider one way to rename them We match them according to their positions in the object vectors which are lexicographically ordered The names of the variables thus matter Thus for the queries a and x man y woman and b a man b woman we would rename lt a b gt into lt x y gt and thus see that and x man y woman neg x man as well as and x man y woman neg y woman are inconsistent This show that b entails a and vice versa However for b b man a woman neither and x man y woman neg x woman nor and x man y woman neg y man are inconsistent This shows that there is no entailment in this case Note that it is quite reasonable not to consider all possible mappings if one accepted the typed tuple positions perspective Under this perspective the object
276. s of the concept mother concept ancestors mother 33 get all sub concepts of the concept man concept descendants man 3 get all transitive roles in the T box family all transitive roles 333 some A box queries 3 Is Doris a woman individual instance doris woman Of which concepts is Eve an instance individual types eve 33 get all descendants of Alice individual fillers alice has descendant individual direct types eve concept instances sister describe individual doris describe individual charles Racer Systems GmbH amp Co KG http www racer systems com Appendix C SWRL Example Ontology lt xml version 1 0 gt lt rdf RDF xmlns xmlns xmlns xsd http www w3 0rg 2001 XMLSchema xmlns xmlns xmlns xmlns swrlb http www w3 org 2003 11 swrlb swrl http www w3 org 2003 11 swrl rdf http www w3 org 1999 02 22 rdf syntax ns rdfs http www w3 org 2000 01 rdf schema owl http www w3 org 2002 07 owl daml http www daml org 2001 03 daml oil xmlns http www owl ontologies com unnamed owl xmlns dc http purl org dc elements 1 1 xml base http www owl ontologies com unnamed owl gt lt owl Ontology rdf about gt lt owl imports rdf resource http www daml org rules proposal swrlb owl gt lt owl imports rdf resource http www daml org rules proposal swrl owl gt lt owl Ontology gt lt owl Class rdf I
277. s which correspond to the representative concepts of the taxonomy will be instances of the virtual concept taxonomy node The edges of the taxonomy are represented using virtual has child edges role assertions In queries also the inverse of has child can be used has parent as well as the transitive closures has descendant and has ancestor In principle the taxonomy ABox could also contain additional edges e g representing disjointness information between concepts This will be supported in a future nRQL version 6 1 7 2 Using TBOX RETRIEVE to Query the Taxonomy We will no illustrate the TBox queries using the above constructed taxonomy First let us retrieve all child concepts subsumed concepts of the concept woman tbox retrieve x woman x has child gt x woman with children x mother The inverse role is called has parent tbox retrieve x x woman has parent gt x mother x woman with children Nodes in the taxonomy ABox can be accessed by name or by concept Racer Systems GmbH amp Co KG http www racer systems com 162 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL N tbox retrieve x x mother x woman with children x mother Vv N tbox retrieve mother mother mother mother mother v The set of representative concepts can be retrieved using the virtual query concept taxonomy node tbox retrieve x x tax
278. sFor x a http www lehigh edu 7Ezhp2 2004 0401 univ bench owl name x a http www lehigh edu 7Ezhp2 2004 0401 univ bench owl emailAddress x a http www lehigh edu 7Ezhp2 2004 0401 univ bench owl telephone x datatype fillers http www lehigh edu 7Ezhp2 2004 0401 univ bench owl name x datatype fillers http www lehigh edu 7Ezhp2 2004 0401 univ bench owl emailAddress x datatype fillers http www lehigh edu 7Ezhp2 2004 0401 univ bench owl telephone x 333 Query 5 benchmark universities departments 5 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Person x http www Department0 University0 edul Inttp www lehigh edu X7Ezhp2 2004 0401 univ bench owl member0f x 333 Query 6 Racer Systems GmbH amp Co KG http www racer systems com 255 benchmark universities departments 6 x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Student x 333 Query 7 benchmark universities departments 7 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Student y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Course http www Department0O University0 edu AssociateProfessor0 y Inttp www lehigh edu 7Ezhp2 2004 0401 univ bench owl teacher0f x y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl takesCourse x y 333 Query 8
279. scriptions e g defined using RDF Last but not least the system can also be used for modal logics such as Km 1 1 Views on RacerPro 1 1 1 RacerPro as a Semantic Web Reasoning System and Information Repository The semantic web is aimed at providing machine understandable web resources or by augmenting existing resources with machine understandable meta data An important aspect of future systems exploiting these resources is the ability to process OWL Web Ontology Language documents OWL KBs which is the official semantic web ontology language Ontologies may be taken off the shelf or may be extended for domain specific purposes domain specific ontologies extend core ontologies For doing this a reasoning system is required as part of the ontology editing system RacerPro can process OWL Lite as well as OWL DL documents knowledge bases Some restrictions apply however OWL DL documents are processed with approximations for nominals in class expressions and user defined XML datatypes are not yet supported A first implementation of the semantic web rule language SWRL is provided with RacerPro 1 9 see below for a description of the semantics of rules in this initial version The following services are provided for OWL ontologies and RDF data descriptions e Check the consistency of an OWL ontology and a set of data descriptions Racer Systems GmbH amp Co KG http www racer systems com 2 CHAPTER 1 INTR
280. se note that the same as assertion is also available in OWL 6 1 1 5 Auxiliary Query Atoms Some auxiliary query atoms are offered by nRQL However these atoms are not as important as the atoms we have already discussed 6 1 1 5 1 HAS KNOWN SUCCESSOR Query Atoms Sometimes one just wants to retrieve individuals which have a certain role successor but the actual successor itself is of no interest full reset Racer Systems GmbH amp Co KG http www racer systems com 110 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL gt okay full reset N racer read file nrql user guide examples family racer OKAY Vv retrieve x x has known successor has child x betty x alice Vv This is in fact equivalent to retrieve x x y has child gt x betty x alice However the situation changes if the NAF negated variants of these atoms are considered retrieve x meg x y has child gt x eve x doris x charles x betty x alice retrieve x meg x has known successor has child gt x eve x doris x charles The explanation for the difference is that for the first query the two dimensional comple ment of x y has child is constructed with neg x y has child and this set of pairs is projected to their first components tuple positions In the second query a one dimensional complement
281. set of consequences at a time style nRQL doesn t offer a predefined rule application strategy However API functions are pro vided which enable the client user side definition of different rule application strategies See Section 6 2 4 Racer Systems GmbH amp Co KG http www racer systems com 160 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL 6 1 7 Complex TBox Queries I ope 7 vat gt woman with childrei Pi I 1 bottom s Fi Figure 6 1 Left Taxonomy Right Taxonomy ABox Dashed nodes represent non representative concepts Dashed arrows represent the has synonym relationship the other arrows the has child relationship Complex TBox queries are used to query the taxonomy of a TBox for certain sub superclass relationships The idea is simple whereas the ABox queries we have used so far bind vari ables against ABox individuals the variables in a TBox query are bound against taxonomy nodes Node that a taxonomy node represents a set of synonym atomic concept names 6 1 7 1 The Relational Structure of a TBox Taxonomy The taxonomy of a TBox is a directed acyclic graph DAG A node in this DAG represents an equivalence class of synonym concept names An edge between two nodes represents a direct subsumer of relationship between the corresponding concepts For example the taxonomy produced by the following example is depicted on the left hand side of Figure 6 1
282. sic engine for implementing a very large subset of the OWL QL query language see above 1 1 3 RacerPro as a Combination of Description Logic and Specific Re lational Algebras For some representation purposes e g reasoning about spatial relations such as contains touching etc relational algebras and constraint propagation have proven to be useful in practice RacerPro combines description logics reasoning with for instance reasoning about spatial or temporal relations within the A box query language nRQL Bindings for query variables that are determined by A box reasoning can be further tested with respect to an associated constraint network of spatial or temporal relationships Although RacerPro is one of the first systems supporting this kind of reasoning in com bination with description logics or OWL we expect that international standardization efforts will also cover these important representation constructs in the near future Note also that the semantically well founded treatment can hardly be efficiently achieved using rule systems 1 2 Application Areas and Usage Scenarios Description logic systems are no databases Although one can use RacerPro for storing data using A boxes probably databases provide better support with respect to persistency and transactions However databases can hardly be used if indefinite information is to be handled in an application e g John was seen playing with a ball but I cannot re
283. sistency 188 Racer Systems GmbH amp Co KG http www racer systems com INDEX 265 query inference 188 query life cycle 175 query predicate 153 query realization 199 query realizer 199 query reasoning 188 query repository 195 query subsumption 190 querying qualitative spatial networks 161 range restriction 61 RCC consistency 161 RCC constraint checking 162 RCC query 162 RCC substrate 161 RDF edge query 100 RDF instance query 95 RDF node query 95 referencing a data substrate node 153 region connection calculus 161 relational algebra 4 retraction 67 retrieving concrete domain values 110 role chains in queries 104 role hierarchy 61 role query atom 96 role query atom with features 98 role query atom with negated role 97 rule 139 rule antecedent 139 rule application 139 176 rule application strategy 176 rule consequence 139 rule engine 139 172 rule firing 176 rule life cycle 176 rule postcondition 139 rule precondition 139 rule strategy 176 rules and pseudo nominals 142 rules and the concrete domain 141 scalable query answering 179 semantic cache 195 semantic web 1 semantically identical individuals 99 Racer Systems GmbH amp Co KG http www racer systems com 266 INDEX semi structured data 149 services 1 3 set at a time mode 172 signature 56 socket interface 20 spatial query 162 spatial reason
284. so be used to address the concrete domain part of a RacerPro ABox This means you can add constrained as well as constraints assertions with a rule The syntax for the constraints which is inherited from the RacerPro API function constrained entailed p has been extended so that also head projection operators are valid as parameters in the constraints The following RacerPorter log demonstrates these possibilities gt full reset okay full reset define concrete domain attribute age type cardinal gt OKAY firerule nil instance new ind a an age gt instance a an age firerule nil instance new ind b an age gt instance b an age firerule and x an age y an age neg same as x y constrained x new ind age of x age constrained y new ind age of y age how many 1 gt constrained a age of a age constrained b age of b age firerule Racer Systems GmbH amp Co KG http www racer systems com 156 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL and a an age b an age constraints age b 30 age a gt constraints age of b 30 age of a retrieve x x age 30 gt NIL constraints age of a 30 gt OKAY retrieve x x age 30 gt x a retrieve x told value age x x age 60 gt x b told values age x 60 firerule
285. spibase E E xsp fractionDigits ve i Noli D 5 Navigator 2 _ co Form Diagram Graph Source Code vo 1E 1 imports 2 Instances Domain SPARQL Rules inferences b Bm 07 One IN EESE V amp Ontologies Ta http protege stanford edu plugins owl protege from Ontologies Common protege owl gt amp Common Y amp Examples a family swrl owl A food rdt 2 geotravel owl I pizzaow A wine rdf gt amp ow s a cwn bd los Figure 2 26 The TopBraid Composer interface for displaying and editing ontologies Racer Systems GmbH amp Co KG http www racer systems com 44 CHAPTER 2 USING RACERPRO 2 6 SWRL Semantic Web Rule Language The application of rules provides for optimized manipulation of structures in particular in server based environments such as RacerPro Standards such as SWRL ensure the necessary consolidation such that industrial work becomes possible Therefore RacerPro has been extended with support for applying SWRL rules to instances mentioned in an OWL ontology or corresponding RDF data descriptions The RacerPro SWRL rule engine is currently being extended to cope with OWL datatypes A first experimental SWRL implementation is part of RacerPro 1 9 8 uncle Prot g 3 1 1 file Users rm Desktop Work SWRL Presentation uncle pprj OWL Files owl 7 EER PE EE N Ji prot g F Properties Forms D gt Individuals i Metadata 8 SWRL Ru
286. ss of which the individual is also an instance Let us create two concepts c and d such that d is a sub concept child concept of c full reset gt okay full reset define concept c some r top OKAY v define concept d and c e OKAY v We can verify that d is indeed a child concept of c using a so called TBox query see below tbox retrieve x c x has child gt x d Let us create two individuals so that i and j are instances of c moreover j is also an instances of d Racer Systems GmbH amp Co KG http www racer systems com 126 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL N related i j r OKAY v related j k r gt OKAY instance j e gt OKAY retrieve x x c gt x j Cx i retrieve x x d gt x j The previous queries demonstrated that both i as well as j are c instances However only i is a direct c instance We can retrieve these direct instances of c as follows retrieve1 x c lambda x if some lambda subclass retrieve x subclass flatten tbox retrievel c subclass has child lambda subclass subclass subclass reject x x x gt x i Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 127 6 1 4 Complex Queries After having discussed the available query atoms and
287. ssion above Please note that defined queries can also be used within TBox queries see below for more information on TBox queries resp tbox retrieve defquery subconcepts of mother x mother x has descendant gt subconcepts of mother tbox retrieve descendant concept descendant concept subconcepts of mother gt descendant concept grandmother descendant concept bottom Please note that defined queries can also be used in rules as well see below for more information on rules 6 1 5 1 Ambiguous Queries Please note that the query retrieve a b a b mother of is syntactically ambiguous since mother of might be a role name as well The same problem occurs for concept query atoms In case the query is ambiguous nRQL will output a warning define primitive role mother of gt OKAY Racer Systems GmbH amp Co KG http www racer systems com 150 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL retrieve x y x y mother of NRQL WARNING ROLE mother of EXISTS IN TBOX family ASSUMING YOU ARE REFERRING TO THE ROLE mother of gt NIL defquery mother x and x woman x y has child gt mother retrieve x x mother NRQL WARNING CONCEPT mother EXISTS IN TBOX family ASSUMING YOU ARE REFERRING TO THE CONCEPT mother gt x alice x betty So nRQL assumes you are referring to the role resp concept then If you re
288. st names This is demonstrated with the following statement whose result is displayed just below describe individual document 7 DOCUMENT 7 ASSERTIONS DOCUMENT 7 COMPUTER SCIENCE DOCUMENT ROLE FILLERS NIL TOLD ATTRIBUTE FILLERS ISBN 7234567 NUMBER OF COPIES SOLD 4000 DIRECT TYPES COMPUTER SCIENCE BEST SELLER COMPUTER SCIENCE DOCUMENT Thus RacerPro has determined that the individual document 7 is also an instance of the concept computer science best seller This is due to the value of the at tribute number of copies sold and the given sufficient conditions for the concept computer science best seller in the T box document ontology Now we have information about seven documents declared in the A box current document all individuals DOCUMENT 1 DOCUMENT 2 DOCUMENT 3 DOCUMENT 4 DOCUMENT 5 DOCUMENT 6 DOCUMENT 7 In order to delete a document from the A box it is possible to use RacerPro s forget facility The instance assertion can be removed from the A box with the following statement forget instance document 3 book Now asking for all individuals reveal that there are only six individuals left all individuals DOCUMENT 1 DOCUMENT 2 DOCUMENT 4 DOCUMENT 5 DOCUMENT 6 DOCUMENT 7 With the next subscription a fourth client is introduced The query is to retrieve the in stances of book RacerPro s answer is given below subscribe client 4 book Racer Systems GmbH amp Co KG
289. sults in a hybrid representation The substrate layer is useful for the representation of semi structured data nRQL offers various types of substrates e the basic data substrate e the mirror data substrate as well as e the RCC substrate We will discuss each substrate briefly More types of substrates will be added in the fu ture e g the database substrate for coupling an ABox with a relational database is in preparation 6 1 8 1 The Data Substrate Similar to an ABox a data substrate is a relational structure which can be viewed as a node and edge labeled directed graph Nodes are named and have an optional description the node label Edges must always have a description the edge label 6 1 8 1 1 Data Substrate Labels These descriptions of the nodes and or edges are called data substrate labels The exact syntax can be found on Page 193 syntax rule lt data substrate label gt A data substrate label is either e a simple data literal or e a list of data literals or e a list of list of data literals Data literals are taken from the host language Common LISP Symbols strings numbers as well as character are supported Such a data substrate label is similar to a boolean formula in Conjunctive Normal Form CNF For example the data substrate label foo bar 123 3 foobar somehow represents the positive boolean formula foo V bar A 123 3 A foobar Note that foo and
290. t states The corresponding API functions are named 1 all rules 2 ready rules equivalent 3 active rules a running rules b sleeping rules 4 terminated rules all equivalent Racer Systems GmbH amp Co KG http www racer systems com 208 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL As for the queries there are various synonym functions please refer to the Reference Man ual Note that there is no facility which checks for applicable rules and applies them automat ically The rules have to be fired by hand However the API function applicable rules returns the rules which can fire 6 2 4 How to Implement Your Own Rule Application Strategy A rule application strategy consists of two things First the applicable rules have to identi fied Then a subset of these rules must be applied fired Finally in case any of these rules produces more than one set of consequences it must be decided which consequences to add to the ABox nRQL only provides a function which identified the applicable rules applicable rules All other just mentioned aspects of a rule application strategy must be implemented by the client user Regarding the third step deciding which rule consequences to add to an ABox nRQL provides an add all as well as a selected add only certain chosen sets of consequences to the ABox functionality In the latter case nRQL must be advised which consequences to add the appropriate co
291. t tronsitive t Chas child porent has descendant has sibling has sister porent has sibling has brother t has sibling Chas gender feature t ls alice betty charles doris eve top all has child person Csome has child top parent some has sibling top or sister brother top all has sibling or sister brother top all has sister some has gender female top all has brother some has gender male person and human some has gender or female male disjoint female male Cimplies Cimplies woman Cand person some has gender female man Cand person some has gender male equivalent parent and person some has child person equivalent mother Cand woman parent equivalent father Cand man parent Cequivalent grandmother Cand mother some has child some has child person equivalent aunt Cand woman some has sibling parent equivalent uncle and man some has sibling parent Ready localhost 8088 family racer Figure 2 2 The RacerEditor interface for RacerPro Lisp Users rm Desktop Examples famil Racer Systems GmbH amp Co KG http www racer systems com 14 CHAPTER 2 USING RACERPRO oee RacerEditor i l l Possible completions gt Cin knowledge base famil CAPI editor pane 2 signature otomic conc Info woman man r Main cle family owl roles has Chas family racer has s Chas Chas Chas individuals
292. t gt lt cd attribute name gt lt constraints assertion gt gt constraints constraint see RacerPro constraints syntax Note that you can also use lt abox head projection operator gt and lt lambda application gt within constraint expression thus the syntax has been extended by allowing also expression such as constraints age b 30 age a or constraints told value age y age y etc lt forget concept assertion gt gt forget concept assertion lt query object gt lt concept expression2 gt lt forget role assertion gt forget role assertion lt query object gt lt query object gt lt role expressioni gt lt forget constrained assertion gt gt forget constrained assertion lt query object gt Racer Systems GmbH amp Co KG http www racer systems com 202 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL lt forget constraint assertion gt gt see lt cd object gt lt cd attribute name gt RacerPro syntax However the extensions mentioned in lt constraints assertion gt cannot be used here Auxiliary syntax rules lt role chain followed by attribute gt gt lt cd attribute name gt lt role expression gt lt cd attribute name gt lt concept expression gt gt lt cd attribute name gt gt lt cd object name gt gt lt role expressioni gt gt Note
293. t to really add some expressive power to nRQL For example it is possible to retrieve the oldest person from family racer as follows How shall we formulate such a query Obviously a person is the oldest person if there is no other person that is older This intuition gives us the following query retrieve x and x person x an age neg project to x and y person y an age x y constraint age age lt gt x alice Please note that the following query does not work since only the variables mentioned in the projection list of project to are shared between queries retrieve x and x person x an age neg project to y and y person y an age x y constraint age age lt gt x alice x doris x eve x betty x charles The reason is that y is not used in the surrounding query and x is not shared either since it is not mentioned in the projection list after project to Using describe query it is to verify that the query body within project to shares no variables with its surrounding query describe query last gt query 3 accurate processed Racer Systems GmbH amp Co KG http www racer systems com 146 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL retrieve x and x person x an age not project to y ano2 and y ano2 person y ano2 an age x ano1 ano3 y
294. t want to use the tab key to complete a partially typed command A command is executed once it is syntactically complete Racer Systems GmbH amp Co KG http www racer systems com 16 CHAPTER 2 USING RACERPRO e ee RacerPorter Shell Cons TBoxes ABoxes Taxo Role Netw Conc Roles Indivi Asse Queri Rules TBox FAMILY ABox SMITH FAMILY Concept Role Individual Query Rule l Simplify Names CINV HAS DESCENDANT CINV HAS CHILD CINV HAS SISTER CENV HAS SIBLING lt HAS BROTHER CINV HAS GENDER HAS DESCENDANT _HAS CHILD HAS BROTHER HAS SIBLING T uns SISTER HAS GENDER Horizontally Vertically Focus Reset ClearLog RacerPro Log Using Profile Localhost to Connect to Localhost 8088 gt OKAY J Using Profile Localhost to Connect to Localhost 8 88 gt 0KAY O v A Figure 2 5 RacerPorter showing the role hierarchy of the T box family G 00e RacerPorter Shell Cons TBox ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox FAMILY ABox SMITH FAMILY Concept Role Individual Query Rule F Simplify Names RacerPro Listener Cconcept subsumes brother uncle T Cconcept ancestors mother CCPARENT WOMAN PERSON TOP TOP CHUMAN Cconcept descendants man CCUNCLE BOTTOM BOTTOM FATHER BROTHER all transitive roles su HAS D
295. tatype fillers datatype property x x some datatype property integer x http www owl ontologies com Ontology1159352693 0wl test datatype fillers http www owl ontologies com Ontology1159352693 0wl datatype property x 123 Vv retrieve x datatype fillers datatype property x x some datatype property and min 100 max 200 gt x http www owl ontologies com Ontology1159352693 owl test datatype fillers http www owl ontologies com Ontology1159352693 owl datatype property x 123 retrieve x annotation datatype fillers annotation datatype property x x an annotation datatype property Racer Systems GmbH amp Co KG http www racer systems com 118 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL gt x http www owl ontologies com Ontology1159352693 owl test annotation datatype fillers http www owl ontologies com Ontology1159352693 owl annotation datatype property x 456 retrieve x annotation datatype fillers annotation datatype property x x some annotation datatype property integer gt x http www owl ontologies com Ontology1159352693 owl test annotation datatype fillers http www owl ontologies com Ontology1159352693 owl annotation datatype property x 456 retrieve x annotation datatype fillers annotation datatype property x x some annotation data
296. te http www owl ontologies com unnamed owl Rule 11 CRELATED http waw owl ontologies com unnamed owl MARY http www owl ontologies com unnamed ow1 OHN http ww owl ontologies com unnamed owl hasUncle 7 gt 0KAY Figure 2 32 Network view indicating the asserted uncle relation between MARY and JOHN Racer Systems GmbH amp Co KG http www racer systems com 50 CHAPTER 2 USING RACERPRO r e608 RacerPorter Shell Cons TBoxes ABox Taxo Role Netw Conc Roles Indiv Asse Quer Rules TBox temp 1 0w1 ABox temp 1 0wl Concept Role Individual MARY Query Rule v Simplify Names Id State Cheap Next Rule Rule 11 ACCURATE READY TO RUN yes no PREPARE ABOX RULE AND x y hasMother y z hasBrother RELATED x z h E gt Al O Ready Running O waiting O Cheapw ExpensiveW Processed O Acc O inace Status NextSetofConsequences Answer Delete Abort Execute Re Execute Re Prepare Refresh ClearLog RacerPro Log 7 SWRL FORWARD CHAINING Execute http ww owl ontologies com unnamed owl Rule 11 RELATED http ww owl ontologies com unnamed owl1 MARY http ww w owl ontologies com unnamed owl JOHN http ww owl ontologies com unnamed owl hasUncle 7 gt 0KAY Figure 2 33 Rules tab in RacerPorter Various semantics have been proposed for rule languages Rules have a body and a head If there exists a
297. tency w r t a T box Is the set of objects described by a concept empty e Concept subsumption w r t a T box Is there a subset relationship between the set of objects described by two concepts e Find all inconsistent concept names mentioned in a T box Inconsistent concept names result from T box axioms and it is very likely that they are the result of modeling errors e Determine the parents and children of a concept w r t a T box The parents of a concept are the most specific concept names mentioned in a T box which subsume the concept The children of a concept are the most general concept names mentioned in a T box that the concept subsumes Considering all concept names in a T box the parent or children relation defines a graph structure which is often referred to as taxonomy Note that some authors use the name taxonomy as a synonym for ontology Note that whenever a concept is needed as an argument for a query not only predefined names are possible instead concept expressions allow for adaptive formulation of queries that have not been anticipated at system construction time If also an A box is given among others the following types of queries are possible e Check the consistency of an A box w r t a T box Are the restrictions given in an A box w r t a T box too strong i e do they contradict each other Other queries are only possible w r t consistent A boxes e Instance testing w r t an A box and a T box Is the o
298. than CWA However in order to be com pliant with the semantics of OWL RacerPro does not apply the unique name assumption by default If you want this use the following statement before asking any queries set unique name assumption t You might want to put this directive into a file init racer and start RacerPro with the following option RacerPro init init racer 4 6 Differences in Expressivity of Query and Concept Lan guage It should be emphasized that the query language of RacerPro nRQL see Chapter 6 is different from the concept language Feature chains and pseudo nominals are supported in queries but not in the concept language In addition specific concrete domain predicates may be used in query expressions e g substring that cannot be supported in the concept language Racer Systems GmbH amp Co KG http www racer systems com 74 CHAPTER 4 DESCRIPTION LOGIC MODELING WITH RACERPRO 4 7 OWL Interface RacerPro can read RDF RDFS and OWL files see the function owl read file and friends described below Information in an RDF file is represented using an A box in such a way that usually triples are represented as related statements i e the subject of a triple is represented as an individual the property as a role and the object is also represented as an individual The property rdf type is treated in a special way Triples with property rdf type are represented as concept assertions RacerPro do
299. the above descriptions Note that this information is available to queries There are three kinds of nodes 1 The label of a node representing an ABox individual e g alice contains the con juncts marker abox object abox individual The node also contains concept membership information e g for alice top mother This information has been extracted from the given concept assertions in the KB 2 The label of a node representing a concrete domain object e g alice name contains the conjuncts marker abox object abox concrete domain object Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 177 3 The label of a node representing a told concrete domain value datatype value e g 80 Alice contains the conjuncts marker abox object abox concrete domain value Moreover the node has the name of its value and also contains its value contains itself in the label The related constrained and constraints Abox assertions give rise to appropriately labeled edges in the substrate connecting the different types of substrate nodes Let us inspect some edges get data edge description alice betty gt from node alice to node betty edge label has child abox relationship abox role relationship has descendant get data edge description alice alice age gt from node alice to node alice age edge label abox relationship
300. the error message in the console window The option h prints some information about the usage of the program RacerPro RacerPro p h The f RacerPro option has the following meaning The input file is family racer and the queries file is family queries lisp The output of RacerPro is printed into the shell If output is to be printed into a file specify the file with the option o as in the following example RacerPro f family racer q family queries lisp o ouput text Or use the following under Windows as explained above RacerPro c f family racer q family queries lisp o ouput text The syntax for processing input files is determined by RacerPro using the file type file extension If lisp krss or racer is specified a KRSS based syntax is used Other possibilities are rdfs owl and dig If the input file has one of these extensions the respective syntax for the input is assumed The syntax can be enforced with corresponding options instead of f rdfs owl and dig The option xml lt filename gt is provided for historical reasons The input syntax is the older XML syntax for description logic systems This syntax was developed for the FaCT system 7 In the RacerPro server output for query results based on this syntax is also printed using an XML based syntax However the old XML syntax of FaCT is now superseded by the DIG standard Therefore the RacerPro option xml may no longer be supported once the fi
301. this chapter of the user guide we describe an expressive A Box query language for Racer Pro called nRQL nRQL is an acronym for new Racer Query Language pronounce Nercle Previous versions have been called RQL in order to avoid confusion with an RDF query language we have changed the name nRQL can be used to e query RacerPro ABoxes e query RacerPro TBoxes e query RDF S documents e query OWL documents Thus nRQL is e an expressive ABox query language for the very expressive DL ALCOHTR DT e an RDF S query language e an OWL query language nRQL allows for the formulation of so called grounded conjunctive queries A query con tains variables which are bound to ABox OWL RDF S individuals concepts or so called substrate nodes Thus we distinguish between ABox variables TBox variables and substrate variables In case of an ABox variable the variable is bound to those ABox indi viduals that satisfy the query body Within a query body concept and role terms are used to specify retrieval conditions on the bindings of a query variable e g retrieve only those x which are instances of class woman which have a child The features of the nRQL language can be summarized as follows Racer Systems GmbH amp Co KG http www racer systems com 88 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL e nRQL has a well defined syntax and semantics Please refer to the list of research publications at http
302. tly declared its type is implicitly derived from its use in a T box A box An attribute and its type can be declared with the signature form see above or by using the KRSS like form define concrete domain attribute If an attribute is declared to be of type complex is can be used in linear in equations However if an attribute is declare to be of type real or integer it is an error to use this attribute in terms for nonlinear polynoms In a similar way currently an attribute of type integer may not be used in a term for a linear polynoms either If the coefficients are integers then cardinal natural number including 0 for the type of attributes may be used in a linear polynom Furthermore attributes of type string may not be used on polynoms and non strings may not be used in constraints for strings 3 7 Individual Assertions and A boxes An A box contains assertions about individuals The set of individual names or individuals for brevity Z is the signature of the A box The set of individuals must be disjoint to the set of concept names and the set of role names There are four kinds of assertions Concept assertions with instance state that an individual IN is an instance of a spec ified concept C Role assertions with related state that an individual IN is a role filler for a role R with respect to an individual IN 9 Attribute assertions with constrained state that an object ON is a filler for a role R with respect to an indi
303. toTopping HamTopping MozzarellaTop Fiorentina f NamedPizza NECESSARY PeperonataTopping FourSeasons WV hasTopping TomatoTopping HamTopping MozzarellaTopping u Anchovie E AnchoviesTopping FruttiDiMare 3 hasTopping PeperonataTopping CaperTopping Giardiniera 3 hasTopping AnchoviesTopping OliveTopping LaReine 3 hasTopping CaperTopping HamTopping Margherita 3 hasTopping OliveTopping Gea MozzarellaTopping 3 hasTopping HamTopping Del TomatoTopping isTopping MozzarellaTopping Mushroom 3 hasTopping MozzarellaToppi japoletana asTopping TomatoTopping Napolet 3 h Ti Parmense INHERITED PolloAdAstra 2 hasBase PizzaBase from Pizza CE GD Dis a al PrinceCarlo FourSeasons QuattroFormaggi QuattroFormagai AmericanHot 0 Bion Mushroom Siciliana s Parmense SloppyGiuseppe n Napoletana gt PrinceCarlo x n Gaj TE si b Logic View Properties View Figure 2 20 The pizza ontology loaded in Prot g pizza owl Prot g 3 1 1 file Applications Protege_3 1 examples pizza pizza owl ppri OWL Files owl or rdf Y ajea x lt gt SCE ae Properties Forms For Project pizza owl Asserted Hierarchy owl Thing Y DomainConcept Country iceCream v Pi
304. toms work as well for datatype properties but some auxiliary role as sertions must be added nRQL must be put into a special mode as the following example demonstrates full reset okay full reset v N add role assertions for datatype properties v okay adding role assertions for datatype properties J owl read file nrql user guide examples owl ex3 0w1 Vv home mi wessel nrql user guide examples owl ex3 owl retrieve x y x y constraint age age lt x http www owl ontologies com unnamed owl betty y http www owl ontologies com unnamed owl alice v Again complex predicates can be used retrieve x y x y constraint age age lt age 1 30 age 2 gt NIL retrieve x y x y constraint age age lt age 1 29 age 2 gt x http www owl ontologies com unnamed owl betty y http www owl ontologies com unnamed owl alice 6 1 1 4 SAME AS Query Atoms and Synonym Individuals nRQL supports different notions of equality or sameness The first notion is the syntactic notion of name equality Using the binary same as query atom in a complex query it is Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 107 possible to specify that two variables must be bound to the same individual or must be bound to individuals with different names Please note
305. tp www owl ontologies com unnamed owl betty v Again is bound to the current namespace prefix get namespace prefix gt http www owl ontologies com unnamed owl retrieve x y x y http www owl ontologies com unnamed owl hasChild gt x http www owl ontologies com unnamed owl alice y http www owl ontologies com unnamed owl betty Moreover also works for individual names retrieve x x betty hasChild gt x http www owl ontologies com unnamed owl alice Please note that for older version of RacerPro lt 1 9 0 the abbreviation will not work and moreover bars must be used as follows Racer Systems GmbH amp Co KG http www racer systems com 102 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL retrieve x y x y http www owl ontologies com unnamed owl hasChild gt x http www owl ontologies com unnamed owl alice y http www owl ontologies com unnamed owl betty Note that nRQL also provides facilities to query for the fillers of OWL annotation and OWL datatype properties see below 6 1 1 2 4 Explicit and Implicit Role Fillers It is important to understand that only explicitly modeled role fillers can be retrieved with nRQL s role query atoms However some KBs may have implicit logically implied role fillers whose presence is enforced in the logical models of the KB Let us consider family racer again to illustrate
306. ttp www racer systems com 6 1 THE NRQL LANGUAGE 195 The syntax of the head projection operators lt abox head projection operator gt lt attribute name gt lt query object gt told value told values lt attribute name gt lt query object gt told value if exists told values if exists lt attribute name gt lt query object gt datatype filler datatype fillers lt OWL datatype property gt lt query object gt annotation datatype filler annotation datatype fillers lt OWL annotation datatype property gt lt query object gt all types types instantiators all instantiators all types flat types flat all instantiators flat instantiators flat direct types most specific types most specific instantiators direct instantiators direct types flat most specific types flat most specific instantiators flat direct instantiators flat describe individual synonyms lt query object gt lt tbox head projection operator gt concept synonyms describe lt tbox query object gt Query bodies and atoms are defined as follows lt abox query body gt gt lt empty query body gt lt abox query atom gt lt substrate query atom gt nen project to project pi lt def query head gt lt abox query body gt
307. type property and min 300 max 500 gt NIL Note no reasoning for annotation datatype properties Note that we can neither retrieve the RDF S comment annotation Demo Ontology of the whole ontology nor the class annotation This is a test class in that way since nRQL is an ABox query language and there are simply no corresponding ABox individuals representing the whole ontology and or the classes However the so called mirror data substrate will among other things contain nodes representing the used XSD literals as well as nodes representing the individual OWL classes and thus using a mirror data substrate instead of an ABox the annotation fillers of classes can be retrieved see below 6 1 3 Lambda Head Operators to Evaluate Expressions The application of a head projection operator can be understood as a function applica tion So called lambda expressions can denote anonymous functions nRQL allows for the specification of lambda expressions in the head of a query nRQL uses a Lisp dialect which we call MiniLisp MiniLisp is a termination safe expression language not a programming language For example consider an ABox representing material objects in the physical world having width and length and we want to computed and return the area of these objects with a query see file lambda ex1 ow1 define concrete domain attribute width type integer gt OKAY define concrete domain attribute length
308. type property relationship owl datatype annotation property relationship http www w3 0rg 2000 01 rdf schema comment from node http www owl ontologies com Ontology1159352693 0wl test to node 456 edge label owl relationship owl datatype property relationship owl datatype annotation property relationship http www owl ontologies com Ontology1159352693 owl annotation datatype property from node http www owl ontologies com Ontology1159352693 0wl test to node Comment for individual test edge label owl relationship owl datatype property relationship owl datatype annotation property relationship http www w3 0rg 2000 01 rdf schema comment It is now easy to retrieve the OWL classes which have an annotation since also the classes and its annotation are represented as nodes in the substrate note that there are no ABox individual though So we can ask for the datatype property annotation of all OWL classes which contain test as a substring retrieve x y and x owl class x y owl datatype annotation property relationship y predicate search test gt x http www owl ontologies com Ontology1159352693 0wl class y This is a test class Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 185 We already know that the prefix always expands into the cur rent default namespace of the O
309. ule consequences can be called to add the chosen sets of consequence manually However add chosen sets of rule consequences cannot be called if the rule is still active A rule which had been aborted never adds rule consequences automatically to an ABox Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 209 Let us consider an example session which demonstrates these aspects of the API full reset okay full reset Vv racer read file home mi wessel nrql user guide examples family no signature racer gt OKAY process tuple at a time okay processing tuple at a time v N describe query processing mode creating substrates of type racer dummy substrate using prefix for injective variables check abox consistency same as is syntactic same as query optimization enabled optimizer uses cardinality heuristics automatically adding rule consequences warnings complete mode mode 3 tuple at a time mode eager deliver kb has changed warning tokens v Please note that automatically adding rule consequences is enabled prepare abox rule x woman instance x mother related x new ind child of x has child gt rule 1 ready to run N ready rules rule 1 v applicable rules rule 1 v get next set of rule consequences rule 1 gt not found execute rule rule 1 gt rule 1 ru
310. ut to check whether individual instance i and C not D returns t The reason is that in this case the index structures for the A box are not recomputed Racer Systems GmbH amp Co KG http www racer systems com 68 CHAPTER 3 RACERPRO KNOWLEDGE BASES Racer Systems GmbH amp Co KG http www racer systems com Chapter 4 Description Logic Modeling with RacerPro There are several excellent books and introductory papers on description logics e g 1 2 In this Chapter we discuss additional issues that are important when RacerPro is used in practical applications 4 1 Representing Data with Description Logics Almost nothing is required to use a description logic inference system to store data In particular there is no need to specify any kind of memory management information as in databases or even in object oriented programming systems In order to make this clear a very simple example is given as follows in knowledge base test instance i a instance i b related i j r Given these declarations for data query functions can be used for data retrieval concept instances and a some r b returns i You could also use the nRQL query language see Chapter 6 retrieve x x and a some a b If you would like to know which fillers actually got names in the A box use the following query retrieve x y and x a x y r y b Racer Systems GmbH amp
311. vely we could also turn on the semantic same as interpretation using the nrql symbols use individual synonym equivalence classes directive retrieve alice or not alice alice nrql equal role not alice grandmother gt alice eve alice doris alice charles alice betty Racer Systems GmbH amp Co KG http www racer systems com 6 1 THE NRQL LANGUAGE 135 Please note that individuals in role query atoms which use the nrql equal role will never be replaced with their representative variables Moreover we can also turn on the semantic same as interpretation as follows use individual synonym equivalence classes gt okay using individual equivalence classes retrieve alice individual synonyms alice alice grandmother dont show lambdas p t alice alice alice grandmother v retrieve alice individual synonyms alice neg alice grandmother dont show lambdas p t gt alice eve eve alice doris doris alice charles charles alice betty betty Note that this rule does not hold for binary atoms see below 6 1 4 3 3 Negation as Failure for Role Query Atoms The neg operator can be applied to role query atoms as well Like for concept query atoms a role query atom and its NAF negated variant are complementary retrieve x y x y has child gt x betty y doris x betty y eve x alice
312. vidual IN Constraints within constraints state relationships between objects of the concrete do main The syntax for constraints is explained in Figure 3 2 Instead of attribute names object names must be used In RacerPro the unique name assumption holds this means that all individual names used in an A box refer to distinct domain objects therefore two names cannot refer to the same domain object Note that the unique name assumption does not hold for object names In the RacerPro system each A box refers to a T box The concept assertions in the A box are interpreted with respect to the concept axioms given in the referenced T box The role assertions are also interpreted according to the role declarations stated in that T box When a new A box is built the T box to be referenced must already exist The same T box may be referred to by several A boxes If no signature is used for the T box the assertions in the Racer Systems GmbH amp Co KG http www racer systems com 66 CHAPTER 3 RACERPRO KNOWLEDGE BASES A box may use new names for roles or concepts which are not mentioned in the T box 3 8 Inference Modes After the declaration of a T box or an A box RacerPro can be instructed to answer queries Processing the knowledge base in order to answer a query may take some time The stan dard inference mode of RacerPro ensures the following behavior Depending on the kind of query RacerPro tries to be as smart as possib
313. with a graphic display RacerPlus To minimize the performance overhead due to network based communica tion between RacerPorter and RacerPro as well as to utilize the computing power offered by a single workstation we introduce RacerPlus an integrated workbench which includes RacerPro and the RacerPorter graphical user interface in a single application Racer Systems GmbH amp Co KG http www racer systems com 6 CHAPTER 1 INTRODUCTION RacerMaster Do you develop complex applications for the Semantic Web Are you re searching in the area of description logic knowledge basis and implement your own systems and algorithms for inference problems using the Common Lisp programming language Then you might be interested in RacerMaster which actually is RacerPro as an object code library fasl file You can develop your own application and use Racer technology without an external server application 1 3 2 Installation The RacerPro system can be obtained from the following web site http www racer systems com A documentation for the RacerPro installation process is included in the installer file that you can download from your personal download page The URL for your personal download page is sent to you via email 1 3 3 Licenses RacerPro is available as a desktop version in which the server and clients run on localhost In addition there are licenses available to run RacerPro on a powerful server machine whereas grap
314. www lehigh edu 7Ezhp2 2004 0401 univ bench owl GraduateStudent x http www Department0O University0 edu GraduateCourse0 http www lehigh edu 7Ezhp2 2004 0401 univ bench owl takesCourse x 333 Query 2 benchmark universities departments 2 and Racer Systems GmbH amp Co KG http www racer systems com 254 APPENDIX D LUBM BENCHMARK x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl GraduateStudent y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl University z http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Department x z http www lehigh edu X7Ezhp2 2004 0401 univ bench owl member0f z y http www lehigh edu X7Ezhp2 2004 0401 univ bench owl subOrganizationOf x y http www lehigh edu 7Ezhp2 2004 0401 univ bench owl undergraduateDegreeFrom x y z 333 Query 3 benchmark universities departments 3 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Publication x http www Department0 University0 edu AssistantProfessor0 http www lehigh edu 7Ezhp2 2004 0401 univ bench owl publicationAuthor x http www Department0O University0 edu AssistantProfessor0 333 Query 4 benchmark universities departments 4 and x http www lehigh edu 7Ezhp2 2004 0401 univ bench owl Professor x http www Department0 University0 edul http www lehigh edu 7Ezhp2 2004 0401 univ bench owl work
315. x resp current tbox of RacerPro Use the buttons Set Racer TBox and Set Racer ABox for this purpose if you really want to change the state of the RacerPro server The current abox and current tbox of the RacerPro server is indicated in the TBoxes resp ABoxes tab by marking the TBoxes resp ABoxes name like this gt gt gt DEFAULT lt lt lt 2 4 2 RacerEditor RacerPorter also includes an Emacs style text editor called RacerEditor which you can use to view edit and process your knowledge bases To open the Editor just click on the New button or push the Open button in the Shell tab and select a text file that should be read into the Editor You can open virtually any file with the Editor from plain text to RDF OWL files and those in RacerPro syntax The editor offers a simple syntax sensitive support by displaying text in different colors and by indicating closing parenthesis At the bottom of the window status lines are displayed that are typical to Emacs They indicate a keystroke combination and give other feedback to user interaction in its first line The bottom line shows more general information the connected RacerPro server if actively connected is shown first As usual for Emacs like editors a modification indicator is shown next means buffer unmodified means buffer modified Then buffer related information follows to the right name of the edited file recognized syntax of the content XML
316. y 2000 Springer Verlag 11 I Horrocks and S Tobies Reasoning with axioms Theory and practice In Proc of the 7th Int Conf on Principles of Knowledge Representation and Reasoning KR 2000 pages 285 296 2000 12 Ian Horrocks and Ulrike Sattler Optimised reasoning for SHZQ In Proc of the 15th Eur Conf on Artificial Intelligence ECAI 2002 pages 277 281 July 2002 13 Ian Horrocks and Ulrike Sattler Decidability of SHZQ with complex role inclusion axioms Artificial Intelligence 160 1 2 79 104 December 2004 14 Ian Horrocks Ulrike Sattler and Stephan Tobies A description logic with transitive and converse roles role hierarchies and qualifying number restrictions LTCS Report 99 08 LuFg Theoretical Computer Science RWTH Aachen Germany 1999 15 Ian Horrocks Ulrike Sattler and Stephan Tobies Practical reasoning for expressive description logics In Harald Ganzinger David McAllester and Andrei Voronkov editors Proceedings of the 6th International Conference on Logic for Programming and Automated Reasoning LPAR 99 number 1705 in Lecture Notes in Computer Science pages 161 180 Springer Verlag 1999 16 C Lutz and M Milicic A tableau algorithm for dls with concrete domains and gcis In Proceedings of the 2005 International Workshop on Description Logics DL2005 number 147 in CEUR WS 2005 17 P F Patel Schneider and B Swartout Description logic knowledge representa tion system specif
317. y can be answered simultaneously concurrently For each query a query answering thread is created Support for different querying modi Set at a time mode In this mode the answer to a query is delivered in one big bunch as a set The nRQL API works in a synchronous fashion This means that the API is blocked not available until the query answer has been computed and delivered Racer Systems GmbH amp Co KG http www racer systems com 91 Tuple at a time mode In this mode the answer to a query is computed and re trieved incrementally tuple by tuple The nRQL API behaves in an asynchronous fashion A client can incrementally request tuples from a query one by one more over while the query is still running another query can be started in parallel This mode comes in two disguises lazy or eager In lazy mode the next tuple is only computed if requested on demand In eager mode additional even if not yet requested tuples are computed in advance in the background The lazy mode maximizes the availability of nRQL whereas the eager mode has the advantage that a request for a next tuple can eventually be satisfied immediately if that tuple has already been computed e The degree of completeness is configurable nRQL offers a complete mode as well as various incomplete modi which differ w r t the degree of completeness they achieve Eventually an incomplete mode will only deliver a subset o
318. y charles x charles y alice Using the neg and project to operators see below it is even possible to retrieve the oldest person the oldest female or oldest male person as follows Racer Systems GmbH amp Co KG http www racer systems com 104 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL retrieve x neg project to x x y constraint age age lt gt x alice retrieve x and x woman neg project to x and y woman x y constraint age age lt x alice v retrieve x and x man neg project to x and y man x y constraint age age lt x charles v 6 1 1 3 1 Role Chains in Constraint Query Atoms Not only attributes can be used in constraint query atoms but role chains ended by an attribute This means that a constraint query atoms such as x y constraint has child age has child age lt can be used to retrieve those not necessarily distinct parents x y such that x has at least one child that is younger than one of y s children Moreover role terms can be used in these role chains thus negated and inverse roles are admissible 6 1 1 3 2 Complex Predicates So far we have only used simple concrete domaine predicate e g lt Complex predicates are admissible as well suppose you want to find out who is at least 40 years older than whom retrieve x y x y const
319. y item gt lt data query item gt lt data query predicate gt gt predicate lt 0 ary data query predicate gt lt 1 ary data query predicate gt lt disjunction of data query predicate items gt gt lt 2 ary data query predicate gt lt 2 ary data query predicate gt lt O ary data query predicate gt gt stringp zerop integerp numberp consp symbolp keywordp rationalp floatp minusp Racer Systems GmbH amp Co KG http www racer systems com 200 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL lt 1 ary data query predicate gt gt string string equal string lt string gt string lessp string not lessp string string not equal string lt string gt string greaterp string not greaterp search lt string data literal gt n n gt lt gt lt lt numeric data literal gt find lt character data literal gt string lessp string not lessp string greaterp lt 2 ary data query predicate gt gt string string string equal string not equal string lt string lt string gt string gt string not greaterp gt lt Usa lt Rule consequences lt rule consequence gt gt lt generalized abox assertion gt lt rule anteced
320. y query repository enabled Vv retrieve x x woman x alice x betty x eve x doris Vv Racer Systems GmbH amp Co KG http www racer systems com 6 2 THE NRQL QUERY PROCESSING ENGINE 235 defquery mother with son x y and x woman x y has child y man gt mother with son show qbox for abox 333 QBOX FOR racer dummy substrate FOR ABOX smith family 333 O master top query 55 Neve laquery 1 335 ___0 master bottom query gt see output on stdout describe all definitions gt defquery mother with son x y and x woman x y has child y man store substrate for abox test substrate image gt done The substrate can now for example be restored on a different RacerPro server restore substrate test substrate image gt smith family all aboxes gt default smith family all individuals gt eve doris charles betty alice all atomic concepts gt top bottom father uncle woman human male sister brother Racer Systems GmbH amp Co KG http www racer systems com 236 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL grandmother man aunt female person mother parent describe all definitions gt defquery mother with son x y and x woman x y has child y man show qbox for abox 333 QBOX FOR racer dummy substrate FOR ABOX smith family
321. you need one new unique concept name for each ABox individual e g you want to add instance betty concept for betty for betty instance doris concept for doris for doris and so on Sometimes these instance assertion resp con cepts are called pseudo nominals Pseudo nominals can be added with a rule as follows please note that we must use nrql user guide examples family no signature racer since the signature mech anism would prevent the creation of new concept names with the rule So family no signature racer is the same KB as family racer but without the signature declaration full reset okay full reset v J racer read file nrql user guide examples family no signature racer OKAY v N firerule x top instance x new concept concept for x instance eve concept for eve instance doris concept for doris instance charles concept for charles instance betty concept for betty instance alice concept for alice v N abox consistent gt t Conceptually new concept is just another syntax for new ind Due to the added instance assertions queries referencing the pseudo nominals can be an swered retrieve x x some has child concept for doris gt x betty 6 1 6 8 OWL Rules Let us demonstrate how to fire rules on OWL KBs We are using the people pets owl KB full reset gt okay full reset owl read file nrql user guide ex
322. ystems com 6 1 THE NRQL LANGUAGE 139 retrieve x y x y constraint has child age has child age gt x betty y betty x alice y alice retrieve x y neg x y constraint has child age has child age gt x eve y eve x eve y doris x eve y charles x eve y betty x eve y alice x doris y eve x doris y doris x doris y charles x doris y betty x doris y alice x charles y eve x charles y doris x charles y charles x charles y betty x charles y alice x betty y eve x betty y doris x betty y charles x betty y alice x alice y eve x alice y doris x alice y charles x alice y betty Thus together we have our 2 23 25 5 5 pairs again But note that as just explained the following query doesn t return 23 pairs retrieve x y x y constraint has child age has child age lt gt gt x betty y betty x betty y alice x alice y betty x alice y alice 6 1 4 3 5 Negated SAME AS Query Atoms NAF A negated SAME AS query atom simply enumerates the complement of its positive variant Racer Systems GmbH amp Co KG http www racer systems com 140 CHAPTER 6 THE NEW RACERPRO QUERY LANGUAGE NRQL N retrieve x same as x alice
323. zza CheeseyPizza InterestingPizza MeatyPizza Y NamedPizza American AmericanHot Cajun Capricciosa Caprina Fiorentina FourSeasons FruttiDiMare Giardiniera LaReine eg i a iv 0 8 American AmericanHot Cajun ig CajunSpiceTopping Capricciosa I gt Individuals Metadata For Proiect pizza owl e08 For Class Capricciosa Connected to Racer 1 9 0 instance of owl Class Finished Classification complete Annotations s Time to update reasoner 3 222 seconds Time to synchronize 3 546 seconds V Check concept consistency Time to build query 0 0030 seconds Time to send and receive from reasoner 2 609 seconds V Inconsistent concepts IceCrear consistent CheeseyVegetableTopping is inconsistent Time to update Protege OWL 0 102 seconds Y Compute inferred hierarchy Time to build query 0 0020 seconds Time to query reasoner 3 94 seconds Time to update Protege OWL 0 392 seconds v Compute equivalent classes Time to build query 0 0090 seconds Time to query reasoner 0 263 seconds Time to update Protege OWL 0 011 seconds Total time 11 703 seconds Added SpicyTopping Added CheeseyPizza InterestingPizza MeatyPizza Added CheeseyPi amiin man i rdfs label L Value Cap
324. zza SpicyPizza 0 Added InterestingPizza CheeseyPizza SpicyPizzaEquivalent NonVegetarianPizza SpicyPizza Added SpicyTopping Added CheeseyPizza InterestingPizza MeatyPizza Added CheeseyPizza InterestingPizza VegetarianPizzaEquivalent1 VegetarianPizzaEquivalent2 e ae z d Figure 2 22 Inferred information about pizza classes is shown in the lower pane Racer Systems GmbH amp Co KG http www racer systems com 2 5 OTHER GRAPHICAL CLIENT INTERFACES FOR RACERPRO IPizzaToppingl IDomainConceptl TOP TOP Tree Depth 1 he O Tree Graph IVegetableToppingl CiLeekTopping ICaperToppingl I0liveToppingl IRocketToppingl ISpinachToppingi ITomatoToppingl ISpicyToppingl IHotSpicedBeefToppingl lAnchoviesToppingl IPrawnsToppingi IMixedSeafoodToppingl IHamTopping IChickenToppingl IPeperoniSausageToppingl IFishToppingl CIParmaHamToppingl IMeatToppingl CIParmensel lAmericanl IPolloAdAstral iusennel kad O Bottom Horizontally Vertically RacerPro Log Using Profile Localhost to Connect to localhost 8088 gt OKAY 0 Focus Reset ClearLog Concept Query Coherent Classify Sweet Peg IP Ti l CiPepperToppingt Ci JalLapenc Ci Peperonc ClArtichokeTopping CiPetitPoisTopping 41 le 0 RacerPorter Shell Cons TBoxes ABoxes Taxo Role

RacerPro User's Guide Version 1.9.2

Contents

Download Pdf Manuals

Related Search

Related Contents