years ago i got my hands wet with some rudimentary text classification and grouping techniques. it was when google news came out, and it was fluffier than i like my news to be (it has since been tuned a lot better). so i wrote my own.

i wrote up my (admittedly naive and simplistic - but effective) methods here:

http://www.informit.com/articles/printerfriendly.aspx?p=3988...