Making bayesian filter detect some new spam

Peter Bonivart bonivart at opencsw.org
Thu Jun 24 16:13:37 IST 2010


On Thu, Jun 24, 2010 at 4:45 PM, Scott Silva <ssilva at sgvwater.com> wrote:
> It hit on bayes_50. Your system has that score set to 0.00

Isn't it reasonable to score that as zero? Bayes thinks it's 50%
chance it's spam (or 50% chance it's ham). To me that's worth 0.00.

-- 
/peter


More information about the MailScanner mailing list