Computer Science final year project Problem writeup

[Shicky]

Hello all, have to write my problem specification for my final year project and well im a little confused as to how I should structure this. The information I have been given is:

Once you have been allocated a project, you should spend approximately three weeks developing the problem specification, which should give a good overview of the problem to be solved, from the user’s perspective. It should provide a general description of the problem including, as appropriate, the way the problem is solved at present and the advantages and disadvantages of the present method. The goals and requirements should also be stated clearly. The constraints of the proposed solution should be specified, including a description of the environment in which the software is to be developed, i.e. the hardware/software platform on which it will run, performance constraints, other programs to be adapted or interfaced to. Verifiable criteria against which the success of the project is to be judged should be identified. Lastly, this should include a Gantt chart outlining the project plan with major milestones and deliverables highlighted.

So has anyone else completed one of these and if so could I see it? No need for your whole writeup, this is just the initial problem description and like i say could do with being pointed in the right in terms of structure

Thanks

 

[Mr^B]

You are told what the structure should be:

Once you have been allocated a project, you should spend approximately three weeks developing the problem specification, which should give a good overview of the problem to be solved, from the user’s perspective. It should provide a general description of the problem including, as appropriate, the way the problem is solved at present and the advantages and disadvantages of the present method. The goals and requirements should also be stated clearly. The constraints of the proposed solution should be specified, including a description of the environment in which the software is to be developed, i.e. the hardware/software platform on which it will run, performance constraints, other programs to be adapted or interfaced to. Verifiable criteria against which the success of the project is to be judged should be identified. Lastly, this should include a Gantt chart outlining the project plan with major milestones and deliverables highlighted.

So it's:

1. General description of the problem
2. The way the problem is solved at present
3. The advantages and disadvantages of the present method
4. The goals and requirements
5. The constraints
6. Criteria against which the success of the project is to be judged
7. Project plan

 

[Shicky]

Yes thank you I have picked out pretty much exactly what you have from that text. But Im still not sure how to approach it in terms of structure, I didnt think it would be a simple document listing those items you picked out. Im curious what others have put down for their projects both past and present

 

[MathewS]

Have you asked your project supervisor for any past dissertations, that you can look at to see how previous students have done theirs?
They don't even need to be related to your project, just seeing how others have been structured can be of help.

 

[md85]

My old uni (Cardiff) seem to still have their final year project guide up, which I seem to remember finding quite useful for structuring and so on.

http://www.cs.cf.ac.uk/currentstudents/project-report.pdf

 

[Mr^B]

Yes, but this is a problem specification, not a complete project write-up. There's no need to overcomplicate it, just deliver what has been asked for.

What word-limit do you have? As that will drive the structure to a degree. If it's quite small, then I'm guessing a paragraph for each section is enough. I'd certainly struggle to write more than a para on each of those sections without resorting to the usual sort of waffle you find padding out most project documentation.

 

[AJK]

Word limit is a good guide - the word limit for my final year project specification was, if I remember correctly, just 2000. This is quite (very?) easy to reach assuming that you cover the 7 points Mr^B posted above.

In terms of actual layout of the document, a subheading and paragraph for each point is probably enough. Where you feel you have more relevant information to include, or where a diagram or similar would be useful, just include that information in an appropriate way. (But don't pad for the sake of padding!)

Importantly; don't overthink it. The specification is the easy part - you are basically just stating what you are going to do, and how you will go about it. Wait til you get to the final write-up!

 

[ZombieFan]

What I generally do for most documents (including specs) is:

1. Add a centred title at the top
2. Add a (automatic) contents page
3. Add each of the 7 items listed above as left-aligned sub title. Add page breaks before each title.
4. Jot notes down in each section about what you want it to contain. Don't worry about complete sentences at the moment. Just general notes, bullet points etc.
5. Expand the notes you've added into complete sentences. Add sub-titles where required (usually as a result of the bullet points).

Before you know it, you've got yourself a half-decent document Then just revise little bits and pieces as required.

 

[NickK]

1. General description of the problem
2. The way the problem is solved at present
3. The advantages and disadvantages of the present method
4. The goals and requirements
5. The constraints
6. Criteria against which the success of the project is to be judged
7. Project plan

I think the point "user’s perspective" is important too.

 

[Mr^B]

I think the point "user’s perspective" is important too.

I think that's implied - I have no other idea whose perspective it would be from...

 

[tntcoder]

Also I would recommend using LaTeX to do it.

 

[AJK]

Also I would recommend using LaTeX to do it.

God, no, really? :-/

(I didn't get on with it.)

 

[tntcoder]

God, no, really? :-/

(I didn't get on with it.)

It's pretty standard in decent Computer Science departments and well respected. Does take a while to get to grips with though as you say. Not essential but it certainly looks better than MS word and is more suitable uni work imo.

 

[Phil99]

God, no, really? :-/

(I didn't get on with it.)

Was there any particular part that you didn't get on with? If you can do a Computer Science course, you can do LaTeX! It makes life sooo much easier when writing reports

 

[AJK]

Phil: true in theory, but I remember that Word seemed to provide the required indexing and referencing features that I needed at the time; LaTeX seemed like an unnecessary extra item to get to grips with in a busy final year. Must admit though, once I'd made that initial decision I didn't spend a lot of time looking into it, so I'm quite prepared to believe it's not that compex really

 

[robmiller]

If LaTeX is too much you should at least use LyX or a similar WYSIWYM frontend. I wrote an undergrad humanities dissertation in pure LaTeX, I don't think you have any excuse if you're doing CS!

 

[Phil99]

Phil: true in theory, but I remember that Word seemed to provide the required indexing and referencing features that I needed at the time; LaTeX seemed like an unnecessary extra item to get to grips with in a busy final year. Must admit though, once I'd made that initial decision I didn't spend a lot of time looking into it, so I'm quite prepared to believe it's not that compex really

Yeah, it's one of those things where it initially seems more complicated as it can look a little daunting to learn, but once you've had an hour with it you wouldn't believe how you ever coped with MS Word

 

[Phreaky]

LaTeX is useful for some things and completely overkill for others. When I did my CS dissertation I did it in Word and got along fine (and got 1st ), but a friend who is doing a PhD swears by it.

 

[StephenLawUK]

Ewww Latex had to use that last year in Computer Science at Surrey. It has its uses but is complete overkill and effert for a plain text document with some images in it. Only time it is useful is really for any paper that requires mathematical formulas and you need to convert it to PDF as Word messes that stuff up when u convert to PDF leaving gaps in your PDF.

As for stucture for report writing im a fan of writing in conferance paper format. Writing in a paper in a formal standard gives it a really professional look and will make it stand out to your project supervisor especially if everyone else is writing a bog standard word doc with headings. Im a fan of IEEE myself you can get the template here in Word and Latex formats.

http://www.ieee.org/conferences_events/conferences/publishing/templates.html