You are viewing a plain text version of this content. The canonical link for it is here.

Posted to users@spamassassin.apache.org by Justin Mason <jm...@jmason.org> on 2004/06/08 07:45:37 UTC

Re: OT: Comparisons of different Bayes Engines, etc.

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Matt Kettler writes:
> Here's some real tests for a more mathematical approach:
> http://freshmeat.net/articles/view/964/

Actually, this one's not much use either -- the author didn't use
SpamAssassin's Bayes for some reason, but compared it against several
hand-trained Bayesian filters.  Not good methodology ;)

- --j.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
Comment: Exmh CVS

iD8DBQFAxVKBQTcbUG5Y7woRAv4fAKCNKIGx8O8Mk4FiPjPBYOwGiLd+SgCfSrDV
mA0Y4+JOKYsCCMoNxCYudeY=
=PXM/
-----END PGP SIGNATURE-----

Re: OT: Comparisons of different Bayes Engines, etc.

Posted by Lucas Albers <ad...@cs.montana.edu>.

Justin Mason said:
> Actually, this one's not much use either -- the author didn't use
> SpamAssassin's Bayes for some reason, but compared it against several
> hand-trained Bayesian filters.  Not good methodology ;)
It will be an unusual comparsion of bayes systems in a few months when 3.0
comes out and people write plugins for crm114/DSPAM/bogofilter and then
they can can included in spamassassin.

err:
Spamassassin now uses all bayesian filters known to man.
They've also added support for the temporal predictor plugin, which
detects the spam before it is sent.

-- 
Luke Computer Science System Administrator
Security Administrator,College of Engineering
Montana State University-Bozeman,Montana