project ideas on algorithms ,data structures and parallel programming

Question

Hi, I am a final year Computer Science and Engineering student.As part of the final year project i was planning to implement a graph algorithms library(using parallelization) for c++ but came to know that they have already been implemented.

I would be very grateful if you people could suggest me some other project ideas related to my interest.My interest areas are :-

1)Data Structures .

2)Parallel Programming.

3)Algorithms(presently only know algorithms given in "Introduction to Algorithms" by CLRS)

4)Text Compression.

I am looking for a topic which will require intensive use of data structures,algorithms and would like to do the implementation in C++/Java.

I did browse through other similar threads here but couldn't find anything of my interest.

Thanks

Answer 1

Implement a search engine?

Answer 2

My serious advice is to carefully tailor your final year project to what will get a good grade. For once in your student live, right at the end, enjoying yourself is secondary. You don't want to regret that you messed it up.

You'll be very lucky if all your tutors and those involved in actually grading your assignments are actually comfortable with those subjects you listed ;-) You don't want your assignment to be hit by a wall of under-appreciation and ignorance.

The right thing to do is to suck up to your tutors and get them to suggest avenues. You might start by saying things like "I'm very interested in text compression, and was wondering if you are interested enough in it to to mentor me through my final year project if I do text compression?" and other horrid sucky things! ;-D

Answer 3

What are some of your higher level interests? The 4 things you describe are tools to get the job done, but not enough to figure out what you really like.

For one of my final projects I implemented marching cubes on arbitrary isosurfaces in PBRT.

If I had access to +4 core hardware (hopefully in the hundreds), I would write my own parallel ray tracer.

Answer 4

My first advice was going to be parallelized graphing algorithms till I saw you'd already looked at that :) I know there has been some research done on parallelizing various parts of graph theory, I've done some myself, but I don't think the field has been exhausted yet.

Often, one way to look for ideas is to read paper on the subjects you are interested in and look for what the authors left out. Is there an area of research they glossed over, special cases in their algorithms that are still uncovered, or did they write a purely theoretical paper neglecting any real world testing and data? All of this offers new avenues for you.

Hope you find something! I don't really know that much on what the hot areas in text compression are, but I'd be interested in knowing what you chose.