This is a (naive) example of naive bayesian text classification.
The perl5 script is derived from code written by John Graham-Cumming
and found in the recent issue of Dr. Dobbs Journal (May 2005) in an
article entitled "Naive Bayesian Text Classification".
The perl6 version is mostly a straight port of that script.
The best way to see what this does is to run the `run.pl` file. It
will run both the perl5 version and the perl6 version one after the
other.
CAVEATS
The 'training' files I have supplied (apples, oranges, grages) are fairly
minimal, and the test file (test) is not so great either. I invite anyone
to add better training and test files to this and/or improve upon the
code itself.