This is a small project that I did for a Machine Learning course. 
It reads in a bunch (a few thousand) small "documents", which are only a few or fewer sentences long.
Then, TF-IDF is performed on them to try to parse out and determine for future articles which words
will probably be the most important ones. The results weren't as useful as I hoped.

The JExcel API was used in this for reading Excel files.

I intend to go back and rewrite much of this for optimization purposes, but, then again,
I've had that intention for quite a while now.

It was a small project for class, I'm not (and wasn't at the time) too worried about it.