I’ll look around to see if I have any of my old Lisp code still. If not, I can at least point you at the algorithms and help you with the (undocumented) file formats.

There’s lots of good news here:
- the datasets are generally “big lists of (small positive) integers” (plus a hash of metadata), which Lisp can slice through like a hot knife through butter
- the state-of-the-art algorithms aren’t that complex (or, alternatively, there’s lots of exploration left to be done!)
- there’s a ton of free data to play with, as you’ve found
- there’s even some free software to be found, though sometimes you have to type it in and port it from ALGOL or a very early FORTRAN and then fix the bugs
- a lot of the algorithms are more or less just “try to do by computer what your eye can do”, so you can make pretty graphics and test your heuristic by just looking at it

Let me know if you have any questions — I’m kind of busy these days, but this sounds like fun.