Thesis Project Management
Contents
1. excuse and make amends 5 Closing remarks Rest me to say that these are only suggestions and guidelines They are based on related material and my own experiences There is not a one way to do a thesis and I have tried different approaches with different theses I completed successfully in the sciences engineering and humanities which partially depended on the topic of the thesis Making a schedule of activities including a prioritisation of activities sticking to them and keeping a logbook invariably were useful and start writing early References 1 M Ashby How to write a paper January 2000 engineering Department University of Cambridge Online Available http www mech eng cam ac uk mmd ashby paper V6 pdf 2 P J Denning D E Comer D Gries M C Mulder A Tucker A J Turner and P R Young Computing as a discipline Communications of the ACM vol 32 no 1 pp 9 23 1989 3 A H Eden Three paradigms of computer science Minds amp Machines vol 17 pp 135 167 2007 4 G D Gopen and J A Swan The science of scientific writing American Scientist Online vol November December 1990 Online Available http www docstyles com library ascience pdf 5 C Johnson What is research in computing science glasgow Interactive Systems Group GIST Department of Computer Science Glasgow University Glasgow G12 8QQ Online Available http2. column An example is shown in Figure 3 For instance it states that in the week of May 10 the statement of requirements is scheduled to be completed and indeed this was the case 11 5 finished revised SoR It also contains text like 17 6 See lab notes on testing today 3 5h which indicates there is a clear distinction between the kind of things you did testing and the technical issues encountered documented in the lab notes the project log is for recording the former information It also contains some tasks and ideas to possibly follow up on 4 9 added table 3 4 on design features Re read correct this later and 30 8 or maybe add an appendix on a brief explanation of ORM respectively 4 3 Review and revise Despite good planning you may not keep up with your planning The way to deal with the delay depends on what causes it There can be many reasons for delays e g 1 You underestimated the problem you aimed to solve because a It is harder than you thought and doing task x takes more time Tasks Comments 1 Statement of Intro chapter with 10 5 made the required changes to the ER model entered meeting answers in doc file and requirements title aim and partially adjusted the SoR to reflect yesterday s meeting 4h 2 Introduction description SoR and 11 5 finished revised SoR started filling Some of the appendices for TMAO 2 added more detail chapter report the relation diagram to the schedu
3. www dcs gla ac uk johnson teaching research_skills research html 6 N Stanton Mastering Communication 3rd ed Macmillan London 1996 Additional sources consulted Nutt Werner How to write a PhD Thesis Heriot Watt University Edinburgh UK 2004 presentation slides The Open University The IT and Computing Project Resources Faculty of Technology version 2 2003 CD ROM CDR0579 11
4. forward calculation means adding your activities sequentially adding estimated workdays estimated begin and end date It will tell you soon enough if you have time left perhaps you can add an activity or short on time even before you have really started The backwards calculation is in that respect more dangerous for your project you may end up with an initial schedule where either you should have started a few weeks ago and are behind already or it gives the false impression you can slack until August If you cannot make a schedule that fits with a realistic amount of estimated workdays for each activity you have a problem There can be several reasons for it Look at it from the bright side it is better to know upfront that your informal intuitive planning does not work so it can be amended early on instead of getting totally stuck or overworked later 4 Avoiding and dealing with delays First and foremost to be able to deal with a delay you need to make yourself aware that a delay exists or when a delay is or can be avoided Keeping a project log and taking the onion approach are two mechanisms for this which will be introduced in the next two sections But despite all things you still may incur a delay for some reason or other suggestions how to cope with that are described in Section 4 1 The onion approach A mechanism to keep you focussed is to use the so called onion approach identify a core of essent
5. methodology you use part of the usual software development cycle may be the identified task to develop a prototype Subtasks can be identified e g determining what features should function in the first prototype e the programming itself e the initial testing e the presentation of the prototype to the employees of Needy Client Corporation and e processing their feedback A milestone in this list of tasks could be the penultimate one it is your first presentable thing of your thesis work to the outside world Observe that here I had not chosen the last event in the list clients complain anyway and once they see a prototype or mock up they tend to generate new requirements and wishes that they want you to take into account which might sound discouraging initially Just that they want more does not imply that what you did is insignificant Milestones for theory oriented activities could be e g the proof the algorithm etc 3 3 Introducing breaks Let s be honest to yourself will you really work on the thesis each day from now until late November If you plan to do so and you skip even a single day you may feel guilty for slacking If you plan and do say on average 30 minutes more per day for two weeks you have a well deserved day off Or perhaps there is a major family event some time in the upcoming months there is no shame in integrating that into your schedule in fact it is Activitie
6. shown to be solved e g a correct proof demonstrating P NP and research questions can be answered Here s a typical example of problematic questions of the form Can X be solved by Y e g Can one use theorem prover ABC to efficiently draw inferences in higher order logics or Can one use algorithm ABC to efficiently compute the solution to the travelling salesman problem It fails if the answer is no and it even fails if the answer is yes but the encoding has not been optimised hence less efficient and when is something efficient compared to what In addition while focussing only on ABC might be sufficient for a BSc project depending on the difficulty of ABC if it has been implemented before etc it is too narrow for a MSc PhD topic A sensible research question for your project could be is ABC faster than DEF in com puting the answer to the travelling salesman problem and if so why if not why not which can be easily reformulated into a hypothesis ABC is faster than DEF in computing the answer to the travelling salesman problem or into a problem it is unknown if ABC faster than DEF in computing the answer to travelling salesman problem although then one has to include in the problem statement that someone cares to know and why The indications provided in this section are not exhaustive but hopefully gives you at least a basic notion To practice f
7. COMP700 Resources Managing your Project j draft v0 3 C Maria Keet email keet ukzn ac za home http www meteck org somputer School o cience University of KwaZulu Nata May 8 2011 Contents 1 Introduction fe ee ee ee ee ee ee ee ee 22 Corea Vile 4 abe thao he ee eee ee oe Oe Oe Deal F POr Oe G6 amp gk oot Ee oe oe ee Ge eh ee et ae ee See nee eee ee ee eee ee eee ee wee ee 2 2 Propos ess Ree YEOH ee ee ew EPRS EENS ai 2 2 4 Core research development 00 000 ee eee 2 2 5 Complete the production of the thesis and defense 3 How to structure your activities 3 1 Milestones and deliverables 0 0 0 0 0 0 00000000 0000008 ea ee ea ee ee er eee eee ee ek eeeaeeeee sete aeeatee eye eaces pen OR aOR ee hPa ee oe ee eee ee ee ee eae ee 4 Avoiding and dealing with delays 4 1 The onion approach cs aoe de eb oa w eee bw a 4 2 Keeping a project log 2 2 ee 4 3 Review and revise aoa a 5 Closing remarks or KRWWNDY WNW OO CO o J Op Oy Co On i jel 1 Introduction Doing your 4th year honours project thesis is quite different from the assignments and exams your are familiar with it spans a longer period and there are no intermediate official checks to test your progress and force you to keep going You will have to rely on your own motivation and planning of your work to complete it To be sure you will have one or two supervisors to guide you with y
8. and do ing research in CS and optionally paradigms in CS 8 In a nutshell there are two principally different thesis be they honours MSc or PhD theses each with sub variants 1 Theory a Novel theoretical advance frameworks theorems proofs b Predominantly literature review Hons MSc only 2 System a Proof of concept implementation b Application through full design cycle Hons only though occasionally seen at MSc level From a CS research perspective the pecking order of importance and coolness and perceived difficulty is 1 a gt 2 a gt 1 b gt 2 b though in countries like the UK and USA which have a more technology oriented culture they probably would switch the position of 1 a and 2 a and let others Europe mainly do the theory development for them From a perspective of someone who aims to get a good job in the IT industry the pecking order of usefulness is probably 2 a gt 2 b gt l a gt 1 b They require different competencies and in part different activities but there are also commonalities which we go through in the next sections 2 2 Core activities There are five main phases to any thesis being 1 Exploration 2 Problem definition 3 Proposal 4 Body of research development including generating a written paper trail 5 Complete the production of the thesis and the defense I will comment on them briefly in the following subsections 2 2 1 Explorat
9. complishment of a task in the project For instance the proof of a theorem or the first version of the implemented algorithm They help you to be pleased satisfied proud of the progress you have made and they will keep you focussed on the important tasks to do You should include both in you schedule 3 2 Dividing up the tasks We now have a list of the general high level tasks that have to be carried out and some basics regarding the deliverables and milestones that have to be included They do not make for a schedule yet What is missing is a further break down into subtasks For instance take literature research This entails sourcing the literature reading it analysing it summarise pros and cons of their proposal follow up on their references making a bibliography and e writing a State of the art section in your thesis that synthesises the relevant literature Task planning does not have to go into extreme detail For instance for making a bibliogra phy planning very fine grained subtasks like explore reference management software read the bibtex manual is a bit over the top in my opinion you can note those in your project log see Section 4 2 that you did them as part of that task though In addition to the subdivision of the larger tasks you will decide where your actual mile stones will be For instance you are developing an application for Needy Client Corporation and depending on the
10. eory oriented theses may seem somewhat more flexible than system oriented but are not The tasks look like More theory amp literature research Design framework method Evaluate framework method Compare your solution with related works or More theory amp literature research Design language algorithm modelling solution Prove correctness of encoding complexity Compare your solution with related works The most important difference however is that with theory oriented theses writing is doing research i e it is an integral part of the tasks whereas this is not the case with system oriented thesis Sure one has to generate a paper trail with the latter type but when say developing the conceptual data model for the application the model is the real outcome of that task not its explanatory text 2 2 5 Complete the production of the thesis and defense You are expected to provide a written and oral account of your project work There is separate documentation that describes the structure of reports and writing style e g L 4 6 so I will not go into those details here While a presentation takes only about a week to prepare writing it clearly and coherently takes a lot longer and always longer than you think it will take Here at UKZN you are expected to write two different documents one shorter article like document about 4 pages double column that exhibits the core of your work and one longer document wi
11. ials that really must be part of your thesis a list of important components and the nice to have ones This helps to ensure you at least will have you core results by the deadline and if there is time remaining you can gradually add extra layers to the core result For instance your thesis revolves around the novel implementation of some algorithm in a fancy way Having it implemented is a task but so are testing on performance developing a nice user interface and it needs a user manual Obviously the implementation of the algorithm itself is a core task and then the second or the third one and writing a manual Two fictitious examples are shown in Figure How does your onion look like Does it match with your initial schedule Did you set the right priorities from the start Perhaps you have to revise your schedule For instance you may have an order of tasks a b c d e and f in the initial schedule because you know upfront you like doing b more than f but in the logical order of things f just have to be done at least before c and possibly also before b Reading this information at the very beginning of the project duration enables or would have enabled you to think about prioritisation of your activities first and devise a sensible planning based on that Monitoring your schedule then can be done through keeping a project log test with user testin J sample diagrams formalise graphical language i
12. ion The principal purpose of the exploration phase is to become acquainted with the area you have chosen This you do by searching for related work and reading up on background material Your supervisor likely will give you some initial hints where to look which provides an easy start for looking up other resources the papers in the references section of that material and when you go to Google scholar and search for the paper it will return that paper and a cited by link When you click on that link you will see who used the paper for something else which may well be something you wanted to do but someone else already did Some journals and publisher already provide you with a This paper has been cited by section on the official page where the paper was published saving you a Google search those lists are also partial however Also searching with related or slightly broader terms can generate interesting results 2 2 2 Problem definition Having read up on your topic and more importantly having analysed critically the papers you have read you may have observed gaps or limitations Or you had a rough idea of the problem already but you have not been able to find any resource that proposes a satisfactory solution These observations help you to come to a precise definition of the problem that has to be solved or the formulation of a hypothesis to be falsified Understanding the problem s is not as easy as it sound
13. le 1h confined to 15 5 working on the appendix diagrammatic representations screenshots of ER ORM and Appendices LogModel Still need to write out the logical schema with the table definitions 1h 21 5 written the logical model i e the domain and table SQL statements 6h week 13 total T2h June 14 1 First draft Should have an 14 6 reminded Brian of my question about references and literature Made first version of the theory section empty database by database called bact2 db 4h Notes after the table 2 start DB then Start to 15 6 added the notes from yesterday s work ihr Need to try out the new create table def implementation consolidate info on written below Organised files properly in subfolders Read about the Gene Ontology theory chapter Consortium 1h include discussion on 17 6 See lab notes on testing today 3 5h DB modelling 19 6 reading over TMAOS requirements project guidelines and adding stuff to the report outline pretty much dome with O1 of TMAOS still need to read it again at a later date 1 5hr week 18 total 10h Aug 9 Data entry 30 8 re created the logical model due to aforementioned food changes bactdb2 ML still need brushing up to print that out 0 5h reading up on another applied bioinformatics thesis that could be of use to report beef up the techie part of the report ih Or maybe add an appendix on a brief explanation of ORM 7k words is just not enough Made cha
14. mplement Fe fofe aitalan Figure 2 Two fictitious examples of how an onion of tasks may look like for a thesis 4 2 Keeping a project log A project log is a document where you summarise the activities and ideas relevant to the project you did that day and the time spent on it Although it may seem to bit of a tedious task to do at the beginning it is useful for several reasons as you go along e You can see for yourself whether you really spend a reasonable amount of time on the thesis over the weeks and months e You can cross check it with your planning are you working on the tasks you set out to do or being side tracked with other activities e You have a single location where you can write down ideas you may have to follow up which are otherwise easily forgotten about a few months later e If you need to backtrack from a dead end in the development you know where to backtrack to e It provides a trail with which you can demonstrate that it really is your work e It helps you evaluating your work the progress in the development of ideas software and thesis writing and reflect on you work habits were they effective if yes why if not why not Would you do it the same way if you had to do it again Assuming you have an outline of the activities and made some schedule see Section 3 then you can format that into a project log by simply adding a comments and or progress column and a actual time spent
15. nges to bact2 MO better reading of the fact types printout again added changed a few bits in the report reading more of the thesis may be a waste of time re report improvement after all 1 5h 1 9 improved summary last section discussion and the odd footnote 1 5h Added some on usability and evaluation 2h 3 9 reading adding fig 1 1 about ontologies to the report 3h a little time did not record 4 9 added table 3 4 on design features Re read correct this later Added a little to fig 1 1 2 5h week 29 total 12 hrs Figure 3 A section of a project log b You don t have sufficient competencies in topic x or c Scratching the surface revealed a plethora of underlying problems that also need to be solved 2 You had a few drafts but the computer crashed and you did not make a backup 3 You need resource x to continue that was supposed to be available by week 5 but only came available by week 10 4 Family emergency Of ax Delays caused by 1 a are easy to detect if you keep a project log you planned for task x to take a week and after say 4 weeks your still working on it Among other typical causes testing software and bug fixing is a common task that takes more time than one initially anticipates and designing good and efficient algorithms is not as easy as it sounds If you do not have sufficient competencies then you will have to put in extra time and effort consulting textbooks etc If that fail
16. ormulating a problem or research question you could first have a look at how others did it by e g analysing the papers you read finding the formulation of the problem question it should be written somewhere in the introduction or else extracting it from the text 2 2 3 Proposal With the problem definition in hand you can start setting out how a solution should look like and how that is to be done This is formulated into a thesis proposal The purpose of a proposal is to clarify the goal of your thesis it outlines possible steps to carry it out it gives you a yardstick to measure your achievements and it is a first seed text for the thesis It is an important document because if your proposal is good you may well fly through your thesis and obtain a good mark On the other hand if the proposal is bad you may embark a journey on the wrong road When is a proposal not good There can be many reasons for that among others You did not search good enough and the problem has been solved already e The problem is too big to tackle in the short time available to complete your thesis e The problem is too small and narrow for a thesis e There is no clearly measurable outcome planned Nobody cares about solving that problem Interesting and hot topic but there are many well funded research groups working on it already you are most likely going to be scooped and or they will outperform you What makes for a good
17. our project but they are not there to do the work for you In addition the amount of work that is expected from you cannot be done the evening before it like cramming for an exam session but should be spread out over the whole period allotted to the project This will require some planning of the activities you will carry out for the thesis so as to stay on track and stick to the commitment you made to yourself This document provides some suggestions and guidelines how to plan and manage your activities that will help you bring it to successful completion without stressing too much 2 The kind of activities The minimum milestones for your honours project are the presentation of your initial project proposal and the 2 3 pages you will submit the honours mini conference day with a short write up and a presentation of your work and the final thesis For a MSc and PhD thesis these may include an additional year report and presentation and or entry exam and or publications of your results as requirements to graduate But what should you do for it and how to go about doing it This section helps you identifying the subtasks and the next section provides suggestions how to structure them 2 1 Types of theses The kind of activities will vary depending on the nature of your project In the following I speak for CS as discipline only You should first consult the research methods resources about what is CS 2
18. proposal This is not simply the opposite of a bad proposal Here are several parameters The solution answer should serve some purpose Results will be concrete Permits theoretical and practical approaches Feasible within the allotted time and available resources There is faculty competent in the topic who can guide you or they know someone who is and can You are expected to present your proposal to the faculty before embarking on the core tasks of the project precisely to avoid the former and aim for a good proposal 2 2 4 Core research development The tasks to carry out i e the item Body of research development including generating a written paper trail listed above differ along the type of thesis For theory the outcome is always and only in writing for systems development it is both in writing and well documented software Moreover software development has a clear design procedure which should be reflected in the activities you plan For proof of concept software the high level tasks would be something along the line of e More theory amp literature research e Design of the algorithm e Program the algorithm e Test and evaluate the algorithm On the other hand tasks for straight forward application development in say the waterfall methodology would be e Requirements analysis e Design conceptual model and logical model e Program the implementation e Test it e Evaluate with users Th
19. s Info Title Given Fla Prede 2011 man Planned Stz 03 04 05 06 07 08 09 10 11 sampierrojerrt 0 VSampleProject Writing thesis 1 YWriting thesis 50d thesis outline 2 thesis outline 1d write introduction 1 Sw i3 write introduction Sd Ga write M amp M 4 write M amp M 3d results 5 results 15d conclusions and summary 6 conclusions and summan 5d layout 7 layout 2d first draft 8 first draft final thesis 9 final thesis Pepe amp proposal accepted 1C proposal accepted 11 de problem definition and proposal 11 problem definition and propc 4d Do the real work 12 YDo the real work 120d 10 Conceptual modelling 13 Conceptual modelling 30d 14 Design 35d 15 CM approved 15 CM approved 13 16 i 20d 1 Implementation Implementation 4 performance testing Leiw 17 performance testing 5d 16 explore o 1 Vexplore literature research D25 months 1S literature research 25d play with tools 2C play with tools 1d Ga e _ gt lt gt Figure 1 An incomplete Gantt chart made with Merlin better to do so so that you can enjoy the event instead of beating yourself up for not working on the thesis and getting behind in the planning Planning breaks is a bit like fooling yourself but it works wonders So yes you can plan breaks into you schedule I would even say that you are encouraged to do it Why is that so What you have to take in
20. s nor is formulating it properly For instance the following are not problem statements 1 Need to optimise the algorithm for the travelling salesman problem 2 How to support the user in selecting the appropriate part whole relation The first one is not because it indicates a desire to achieve something optimization and the second one is not because it looks at doing something holding his hand as support for no particular reason What in theory could be the real problems underlying those two are mildly exaggerated and simplified 1 Real time applications require calculation of the solution to the travelling salesman prob lem within one minute but all the extant algorithms x y and z take umpteen minutes to compute the answer hence there is no technology that meets the performance require ment demanded by the client 2 Both experienced and novice modellers do not know which part whole relation is appro priate between classes in their domain ontologies and they often choose one arbitrarily which causes an incorrect representation of the knowledge with respect to reality it aims to represent and leads to wrong deductions with respect to the subject domain semantics This in turn leads to erroneous suggestions to users of e g decision support systems in medicine therewith potentially threatening a patient s life Problems are things that can be solved maybe not by you but that is a separate topic and can be
21. s within a reasonable amount of time try harder If that fails you are probably not capable enough in the selected field but perhaps you are in another topic so there might be a way to change topic but this means you will have to put in extra time to make up for the lost time Delays due to 1 c are quite common in science This most likely will result in redefining the topic of your thesis to something more specific to narrow down the problem into something feasible within the allotted time As for cause 2 DO MAKE BACKUPS EACH DAY The third cause is a catch all for various external factors that can mess up your planning Say the Internet was down due to a cyber attack there is a 3 week strike it takes ages to get your test subjects together the specific hardware upgrade hasn t arrived your supervisor decided to go on an extended research visit out of town etc The latter three causes can be anticipated and if you know from your initial planning that that may will occur then revise your schedule such that those things that may take more time are attended first and you can work on other activities in the meantime like reading papers writing a framework of your thesis Your project log will help with that as well to demonstrate you did start early with the possible time consuming things and did all feasible other activities as a stop gap which 10 makes people somewhat more lenient to accept the external factors
22. th the details i e the thesis Upfront it may seem easier to first write the former and then the latter but the quality of the former will be better when you first write the latter and slim it to fit into 4 pages because having the overview and the details allows you to generate a more accurate core Needless to say each of the above mentioned tasks can be broken down into sub tasks down to the point where initially the major task do an honours thesis has been broken down into manageable activities that are doable in the day to day work one step at a time toward that end goal 3 How to structure your activities The previous section already indicated some sequence in the activities but it did not give any practical hints on how to do that for your project work The first you have to determine are your milestones and deliverables then you break up the major activities into smaller tasks and based on that you can make a schedule for your activities 3 1 Milestones and deliverables Deliverables are the concrete things you have to deliver to someone before or at a specified deadline For instance your written proposal and your thesis are deliverables and perhaps your supervisor asks you to hand in a first full draft of the thesis by a certain date They are requirements and constraints that have to be met in order to have a chance to pass the course A milestone is a significant moment of time with a certain ac
23. to account is that the problem with research and development projects is that you always can do more This is in stark contrast with a normal course module where there is a clearly delimited piece of knowledge and skills you have to acquire and once you know it it s enough For theses there is no upper bound really As a consequence you may end up working all the time on it However occasionally taking a step back taking a break and coming back with a fresh mind and energy helps you see things in a new light gets you over a lapse in motivation or makes writer s block vanish Or you can use it as a carrot for yourself to get a less pleasant task done more quickly if I complete x I can go to the beach for a whole weekend and actually enjoy the time off 3 4 Making a schedule We are now ready to start making a schedule using the deliverables milestones main and subtasks There are strict deadlines for the presentations and for the deliverables you have to hand in but the rest is up to you to decide in collaboration with your supervisor In theory there are two ways to put actual dates to the activities calculating forward and calculating backwards the end product should be the same schedule anyway While it is possible to put your schedule in a MS Word or Excel sheet this is rather inflexible to manage because with every little change say you were ill for a week you will have to update all of it manuall
24. y There is not only a plethora of software for scheduling there are also different scheduling systems PERT program evaluation and review techniques and Gantt are the most well known ones Figure 1 shows an incomplete Gantt chart The key elements of such a chart are e Break down of activities and sub activities on the left e Milestones noted on the left visualised with a blue diamond on the right e The estimated workdays for each sub activity 4th column in the left hand pane e The assignment of start and end dates together with the estimated workdays which generates the blue rectangles e The dependencies between the activities Precedes in the left pane visualised as lines connecting diamond and or rectangles in the right hand pane The software signals any inconsistencies For instance if you were to have included a thesis submission deadline of Dec 1 2011 and it being the end of the whole project but subsequently you add an end date for say performance testing to be Dec 5 then it will flag that your schedule won t work Or when you add an activity write conclusions and abstract that you think will take you 5 days 1 day 8 hours to complete and set the start date Nov 24 2011 but your predefined actual days in the week have only 4 hours it will tell you that you go over the project deadline and are supposed to fiddle with the dates and workdays until it fits with its internal constraints Hence the
Download Pdf Manuals
Related Search