Vector space comparisons, like I think the doc BrowserUK pointed to shows, are pretty easy. I played with them a bit (2 years back) and got really cool and easily tweaked results. I learned about it and got going from a nice article on Perl.com, "Building a Vector Space Search Engine in Perl."

I really hope you'll keep us updated if you make progress into this. I was plagiarized at school once and it really stung. Go get 'em.