On Thu, 21 Aug 2003, Todd Vierling wrote:

> On Thu, 21 Aug 2003, Curt Sampson wrote:
>
> : Any python-package gurus out there care to have a go at making a
> : mail/spambayes package (http://spambayes.sourceforge.net/index.html)? It
> : looks to be a rather better bayesian filter than spamprobe.
>
> As a FYI, mail/spamassassin comes with a bayesian subsystem as well.  8-)

Yes, I know, but this one looks to be much, much better. The
tokenization is far more sophisticated, and it's using a different
probability calculation algorithm. As well, though I've not used it,
certain things about the spamassassin filter worry me, e.g.,

    Note that if your mail folders contain things like forwarded spam,
    discussions of spam-catching rules, etc., this will cause trouble. You
    should avoid scanning those messages if possible. (An easy way to do
    this is to move them aside, into a folder which is not scanned.)

For a good Baseyian this would be one of the most important things to scan!

cjs
-- 
Curt Sampson  <cjs@cynic.net>   +81 90 7737 2974   http://www.NetBSD.org
    Don't you know, in this new Dark Age, we're all light.  --XTC