Some Ideas on How to Write Your Project Report

Your report should be 4 to 10 pages, depending on your topic and coverage.
Here are some thoughts; they will apply to your report to some extent
depending on your choices.  Consider this a sample list only.

Let's assume I'm comparing three papers on a similar topic.

A general outline of what you're going to do (in my example, I will
compare and contrast three papers that appeared over two years, 
solving different variants of a problem.)  This is an imaginary example.
At any rate, I explain that this is what I will do.  I might explain why
I though this makes a good topic.

For each paper, include: title, date and venue of publication, 
authors.  What area it falls in.  What is the problem they are
trying to solve.  Is it interesting/practically relevant/difficult?
Has it been open for a long time?  Is it a very well known or well studied
problem?  Go into a bit of motivation.  If the three papers are all on 
the same problem, talk about the similaritites and differences between
their versions, justify.

What was known about this problem before this paper?  What is the main result
of this paper?  Does it solve the problem?  What are the implications of this,
for instance, does it make any practical taks easier?  Maybe implications
on theoretical problems?  Perhaps just satisfies our curiosity?  Brings
a lot of money?

What can you say about the techniques used -- do they borrow from 
previous results?  Are the algorithms complicated, what do they do? 
Do they use any data structures or well known techniques?  Any randomization,
approximation, etc?  Justify those.  Are the solutions good enough?  What
are the running times?

Across papers, how does one paper improve upon or give variations on the 
other?  What is their relationship?  Which one is better/more involved/more
useful/more novel?  The later ones may have the faster algorithms, but perhaps
the earlier ones made the big discoveries.  Would you consider these
results nice/groundbreaking/incremental/useless/phenomenal/...?  Why?

Did any of these papers win any awards?  (most of the time the answer will
be no, so you may skip this part)

Are there any evaluations/tests?  Do you agree with how the results were
evaluated?  What can you say about the algorithms based on their performance
in these evaluations?  Are the evaluations fair?  Does one algorithm produce
better results than another in these?

Are the papers easy to read and verify?  Do they give intuition, motivate,
convince?

What would you do to improve these papers?

What is your final three sentence verdict?