final year project

juror · 05-09-2007 11:44am #1

hey, i'm going into my final (4th) year in a computer science degree and we have a project to do, which we must come up with ourselves (proposals due in fairly soon

) and I have been struggling to come up with something decent...

does anyone have any links to sites where there might be some good ideas or have any ideas themselves?

thanks

euph · 05-09-2007 11:47am

Hey, what college are you in?

Did you do any work experience/placement?

I'm just coming to the end of one and a lot of the stuff they do here is to do with web apps, so definitley thinking along those lines

juror · 05-09-2007 11:51am

euph wrote:

Hey, what college are you in?

Did you do any work experience/placement?

I'm just coming to the end of one and a lot of the stuff they do here is to do with web apps, so definitley thinking along those lines

That'd be giving away too much detail!!
Ya did work experience, finishing in the next week, they've given me a couple of ideas here but they're a bit too complex.

Was thinkning of web apps but everyone seems to do them, and i'm better at programming normal applications

Evil Phil · 05-09-2007 11:55am

If you were experienced enough to do a project on anything at all easily what would that project be?

warrenaldo · 05-09-2007 11:55am

I did a compuet game in C++ using OpenGL library for my final year project. Was pretty cool.

Evil Phil · 05-09-2007 12:06pm

You could write an AI that queries students on their interests and then comes up with final year project ideas, or maybe I'm just being sarcastic

Seriously, ask yourself my question above and it will give you a very good starting point. Once you have that answer all you really have to do is come up with a project that fits the bill, is doable by yourself and doable at a level that will get you great marks.

Higgsy · 05-09-2007 12:10pm

Your project could help you get a job, so I would think along the lines of

Think about something in the line with web services.

Or something to do with AJAX or Java and FLEX

Spud83 · 05-09-2007 12:17pm

I found alot of people did games in my year so the standard of the game had to be high to get a good grade. I would suggest picking an area, or technology that very few other people are going to be using in your year. I think this puts you a step ahea straight away as it makes your project interesting and stands out. I done an application using Java 3D for car modifications. I know that I was a step ahead straight away as although the idea wasn't orginal (ripped from many games), using Java 3D certainly got me extra notice. As I think only one other person was using it and they considering it extremily difficult as people had tried in previous years and failed to get a working project at the end.

EDIT: Try pick somthing you are interested in also, as you will be spending many hours, and long days on it being interested in it and not just wanting to get it working because of the grade will help motivate you to put the hours in.

stipey · 05-09-2007 12:23pm

Write an app that, given, a URL will repeatedly check (every 30 seconds or so) for changes and let you know when it has changed.

I wrote a similar app in Java when i was waiting for tickets to an event to come available for online booking. The webpage in question just said "tickets will go on sale here later this week, check back". So I wrote an app that reads the HTML source from the page, then reads it again. If the two were different then the page had obviously changed - so I knew when to check for tickets instead of going back over and over during the day.

It has worked well for me for about 5 or 6 years now - but I kept it nice and simple a text box for the url and start and stop buttons. You could add additional features to beef it up a bit. (I keep meaning to add some of these to my little app - but I just haven't had the time).

Additional features:
* use a configurable parameter that allows the user to alter the frequency with which sites are checked for changes (i.e. don't hard code 30 seconds)
* allows users to store their favourite sites to watch in a database. user can then pick from a favourite or manually enter a URL.
* watch several sites at once (threading)
* when a change is made - generate an email or text message to the user
* when a change is made - display a popup telling the user what site changed, when it was changed, display any new/changed lines, allow the user to click a button which will fire open the page in the machine's default browswer.
* allow the user to choose between remote or local notification. Remote generates text messages or emails, local opens a popup.
* Allow multiple user accounts - so store emails and mobile numbers.
* Maintenance screens to create new users, sites to watch, change frequency of checks etc...
* Log usage - who logged in, what sites they watched, what sites changed.
* Write any errors to the system or application event logs.

Obviously if apps like this already exist in the public domain, it might not be accepted as a valid final year project.

smcelhinney · 05-09-2007 12:39pm

How about (and Im just being topical here) a website that predicts the probability of road deaths. I know it may be sensitive to some people.. But think of the practical application.

So set up a database with events, use some GIS (or arbitrary identifier) information to locate the event, then predict the probability of frequency etc of subsequent road deaths occuring in the same location.

and then sell it to the government, and ensure I get my cut.

Spud83 · 05-09-2007 12:42pm

smcelhinney wrote:

How about (and Im just being topical here) a website that predicts the probability of road deaths. I know it may be sensitive to some people.. But think of the practical application.

So set up a database with events, use some GIS (or arbitrary identifier) information to locate the event, then predict the probability of frequency etc of subsequent road deaths occuring in the same location.

and then sell it to the government, and ensure I get my cut.

A bit Off Topic but most colleges actually hold the copyright of all projects for at least two years, so in fact they would be the ones selling it to the goverment and you wouldn't get crap unless the college decided to bring you in on it.

juror · 05-09-2007 1:33pm

I like your idea stipey, i might just have to think of some way of bringing a new technology that we haven't used before into it. I'll be giving it some thought anyway
Is it difficult enough to have scope of getting a good mark?

reallybad · 05-09-2007 1:42pm

what about an online poker application? It could either be in stand alone or a flash application - complex system but if you have an interest in something like this then its the way to go forward. speaking from experience there are a lot of foreign poker sites after moving to ireland in the past year, if you had an interest in this it would give you a distinct advantage in getting a job in one of these.

my only advice would be to stick to up and coming or industry strong languages eg .NET, JAVA,AJAX etc

Trust me, if you do a good FYP then you will find it a lot easier get a job. I know that from myself and from the people i look to hire

stipey · 05-09-2007 1:50pm

Is it difficult enough to have scope of getting a good mark?

Long time since i was in final year project land so its difficult for me to say.

For my final year project i wrote a Java app that connected into a SAP system, kicked off a script (written in ABAP) to export xml files relating to all system user's, their job roles and their system permissions. The Java app then read these XML files and highlighted any user's who had more/less permissions than they required based on their job role. As final year projects go, I dont think a whole load of people had ever used SAP in their's before - so mine passed the suitability criteria easily enough.

If you were to implement some of the suggestions below you would be covering networking (reading from the url), threading (watching multiple sites at once), databases and then whatever you use to send emails and text messages. Emails will be a piece of cake as most languages should have extensive libraries to do this. To send a text message you will probably have to use a custom API.

I'm sure you could crowbar in other bits and pieces just for the sake of showing how you could use other technologies. I'm sure if you present any semi thought out idea to a lecturer they will be able to tell you if it is suitable or else suggest changes that would make it more suitable.

Spud83 · 05-09-2007 1:55pm

reallybad wrote:

what about an online poker application? It could either be in stand alone or a flash application - complex system but if you have an interest in something like this then its the way to go forward. speaking from experience there are a lot of foreign poker sites after moving to ireland in the past year, if you had an interest in this it would give you a distinct advantage in getting a job in one of these.

my only advice would be to stick to up and coming or industry strong languages eg .NET, JAVA,AJAX etc

Trust me, if you do a good FYP then you will find it a lot easier get a job. I know that from myself and from the people i look to hire

Yes but I know from my year that they wanted you to come with something slightly original or soemthing that you could put a slightly new twist on. Online poker/gambling is really big now so I'm not sure that would qualify in my college at least that would have been questioned.

reallybad · 05-09-2007 2:00pm

alanmurphy83 wrote:

Yes but I know from my year that they wanted you to come with something slightly original or soemthing that you could put a slightly new twist on. Online poker/gambling is really big now so I'm not sure that would qualify in my college at least that would have been questioned.

what about an application that plays and teaches itself poker using AI?

Spud83 · 05-09-2007 2:14pm

reallybad wrote:

what about an application that plays and teaches itself poker using AI?

Possibly but AI is a difficult area to try and get to work as far as I know. How can you prove that its learning etc...

Ginger · 05-09-2007 5:49pm

You know what I would love.. a text speak to normal speak converter.. market it as to how to understand your kids!

Key concepts, would be regex and usability and adapatabilty to input new types of expersions.. Hell if you could build it as a boards plugins, some users would be eternally grateful!

For mine I wrote an automated computer booking system, including the forcing of people of different O/S's when machines were booked.

All done from a web UI. At the time is was advanced, now its a standard thingy

Other ideas

Smart clients on Windows Mobile (you will have to get an XDA or similar

) Suggested idea would be a GPS style system that used the current cell (mast) location integrated to Goggle Maps and possibly linking to one of the mashups that are out there that tell you the locations of things like buses,when they come by, cinemas (to online booking etc) ad infinitum

Would be a nice project suitibly difficult and flashy enough to have fun with..

Smart clients are fun projects to play with

warrenaldo · 06-09-2007 9:57am

i would fully agree with the comment about - it counts towards getting a job in that area. My interviews were all more interested in my project than my degree.

So pick something that you might want to get into in the future. Make it interesting.

tommycahir · 06-09-2007 10:46am

Heres one that was run last year by my previous college lecturer that might be of use to you.

Proposed Title:
Open Source Web-Based Reporting System.
Description:
This tool will have two distinct user communities: end-users and programmers.
End-users will use this tool via a modern web-based GUI to create and execute reports against
existing RDBMSes, such as MySQL and Oracle. The results of executed reports will appear
within a standard web browser. Upon selecting a database to work with, the tool dynamically
determines the tables used within the database, together with the number and type of fields that
comprise each table. This information is then displayed to the end-user to enable the
specification (creation) of a web-based report. Reports can be saved, edited later, and executed
at any time (via a web-based GUI).
Programmers will use this tool via an appropriate programming API to execute previously
created reports within a web development environment. Support for execution via the Common
Gateway Interface (CGI) is required.
A requirement of this project is to publish the resulting system on the SourceForge site and
solicit feedback from the Open Source Community.

Trampas · 06-09-2007 2:12pm

Look at other final year projects and see if you can take them on further.

They should have a section on future work.

You could always email a lecture or two and ask them for suggestions. They always have a list of project suggestions.

damnyanks · 07-09-2007 12:23am

algorithmic sport betting software. Betfair and betdaq both have bet exchanges with published API's. You could write some software to cover spreads and maake bets where by you nearly always come out on top.

I did something like this with betdaq for fun. It's good stuff

oxygen_old · 08-09-2007 9:14pm

I had an idea for a GPS tracking system that an end user could use to track 3 or 4 people via mobile phone GPS systems. However a mate who is a former s'ware dev degree graduate pointed out that there are existing api's and stuff for GPS tracking systems, and that the whole thing could be whipped together in a couple of weeks, thus I would be shot down by the lectures.

Has anyone any suggestion around this idea. Basically I liked the idea of messing around with GPS and mobile phones, and wanted to code in ASP.NET using C sharp, as I think they would be good languages to have on my CV. I just need to make it allot more computer algorithmy and computer sciencey.

constellation · 08-09-2007 11:28pm

oxygen wrote:

I had an idea for a GPS tracking system that an end user could use to track 3 or 4 people via mobile phone GPS systems.

With the trackees' consent, of course...

Raskolnikov · 08-09-2007 11:40pm

My advice would be to not get too ambitious! The core of your project should be based upon technologies that you're already familiar with. So if the vast majority of your experience is with Java, then stick with using Java!

Raskolnikov · 08-09-2007 11:46pm

oxygen wrote:

Has anyone any suggestion around this idea. Basically I liked the idea of messing around with GPS and mobile phones, and wanted to code in ASP.NET using C sharp, as I think they would be good languages to have on my CV. I just need to make it allot more computer algorithmy and computer sciencey.

Route planning!

Are you aware of the Travelling Salesperson Problem? Basically, it's about finding the shortest path between two points. So, if you've a person on a phone and he is on street x and he needs to get to street y, you devise a system that can give him shortest path directions to his destination.

It's topical and combines computer science theory.

oxygen_old · 09-09-2007 9:46am

That’s not a bad idea. Maybe like an app to organise a team of salesmen with several different stops to make during the day. The algorithm could be how to work out which sales man is best to make which stop.

Another idea I’m bouncing around is a robotic arm, the type that could be used in operations, to do things such as hold incisions open for surgeons, or different organs. I think that might be a bit advanced tho.

juror · 09-09-2007 1:25pm

oxygen wrote:

Another idea I’m bouncing around is a robotic arm, the type that could be used in operations, to do things such as hold incisions open for surgeons, or different organs. I think that might be a bit advanced tho.

Maybe just a little out of my league!

[CrimsonGhost] · 09-09-2007 5:09pm

damnyanks wrote:

algorithmic sport betting software. Betfair and betdaq both have bet exchanges with published API's. You could write some software to cover spreads and maake bets where by you nearly always come out on top.

Known as arbitrage betting. A good example is you bet with a south africa place for the spring bucks to win. You be with an irish company for ireland to win. They'll generally give more favourable odds. Whoever wins you end up your initial outlay plus usally < 3% of that profit. Not alot, but you do it often enough and with big enough somes and guaranteed profits come your way. The maths behind it aren't too hard really, it's just finding bets that meet the criteria that is the challenge. You write a program which does it and you can make money selling the bet tips instead of betting yourself.

As for ideas, if the company you had work placement with has suggestions maybe go with them if you like them, esp if they are somewhat research based. Will almost certainly guarantee you a job when you finish college. Worked for me, when straight back to the company after college as my 4th year project related directly to the work they did.

damnyanks · 09-09-2007 10:26pm

[CrimsonGhost] wrote:

Known as arbitrage betting. A good example is you bet with a south africa place for the spring bucks to win. You be with an irish company for ireland to win. They'll generally give more favourable odds. Whoever wins you end up your initial outlay plus usally < 3% of that profit. Not alot, but you do it often enough and with big enough somes and guaranteed profits come your way. The maths behind it aren't too hard really, it's just finding bets that meet the criteria that is the challenge. You write a program which does it and you can make money selling the bet tips instead of betting yourself.

As for ideas, if the company you had work placement with has suggestions maybe go with them if you like them, esp if they are somewhat research based. Will almost certainly guarantee you a job when you finish college. Worked for me, when straight back to the company after college as my 4th year project related directly to the work they did.

Works better with the sport exchanges as you can sell the bets on (Got Argentina on 400 - 1 , sold it at 25 - 1)

final year project

Comments