Am 2002-09-05 22:16:58 +0900 schrieb michael libby:
> > Good, very good so far. Um, what's a "Bayesian" spam filter?
> There I go assuming everyone reads the same stuff I do... check out 
> http://www.paulgraham.com/spam.html for a good explanation of how this 
> works. Examples are in LISP.

I've read this article, too. ;) I've implemented Paul's plan in Ruby a
few weeks ago. My implementation currently uses qmail with maildirs (but
any directory of mail files will do) and procmail to do the real work.
It seems to work pretty good so far.

You can download my ratespam under this URL:

http://www.ping.de/~flori/programs/ruby/ratespam/ratespam-0.0.1.tgz

Perhaps this is of any use to your project.

-- 
Whenever you find yourself on the side of the majority, it is time to
pause and reflect.
  -- Mark Twain