Ask HN: Final Year CS Degree Project

4 points by mistrQ 15 years ago | 3 comments

Hey HN

So this week I need to choose a project for the final year of my Computer Science degree at the University of Edinburgh.

We have been given a list of ~150 pre-approved projects with supervisors and asked to select 4 top choices. The list has some very technical options such as: "Parallel hashing memory vs content addressable memory", "Static and profile-based parallelisation in the LLVM compiler" etc

These are very theory based and while that is computer science, I'd like to do something more practical. Infact if possible I'd like to kill two birds with one stone and come up with a startup idea that I can use for my project. There are some relatively good pre-proposed projects such as combining travel routes for taxis, spelling correction for twitter etc. However I don't want to rely on trying to get one of these projects.

It's quite difficult to propose your own project as you need to find a supervisor to back your idea. However I think i'll put far more hours in if it's a project I genuinely am interested in and that I can possibly turn into a business.

Any suggestions?

ig1 15 years ago |

What are your interests in terms of CS and startups?

Domain specific search is one area that fits nicely into both categories, there are lots of things for which search algorithms are pretty poor at the moment, for example clothes or sofas.

Fraud detection is another interesting area, although getting the data to analyze might be tricky.

Some other specific ideas:

If you're interested in board games you could build an online competitor to Zillions of Games (a windows games engine that allows non-programmer to create games using S-Expressions).

There's probably some interesting things you can do with Google's english language dataset, for example build a tool that can estimate the age of a work based upon the words used.

Apply cryptography to a modern privacy issue, for example see if you can use cryptographic hashes of DNA for identification purposes as opposed to the DNA itself.

Develop a dynamic pricing/yield-management algorithm for an industry that doesn't normally use them.

JonathanWCurd 15 years ago |

What are you interested in?

Maybe someone on HN can steer you towards something if they know more about where your interests lie?

jcsalterego 15 years ago |

Not so much for business ideas but there are always great ideas for the Google Summer of Code: http://code.google.com/soc/