[Spambayes] Still getting horrible results from SpamBayes - Anyadvice?

Peter Beckman beckman at purplecow.com
Sun Sep 14 16:53:17 EDT 2003


On Sun, 14 Sep 2003, Tim Peters wrote:

> [Adam Lasnik]
> > I've now filtered more mail (approximately 1000 ham and spam), and
> > SpamBayes' performance has remained abysmal (only 1 false negative but
> > more than 30% false negatives) :(

 Filtered?  Or trained Spambayes on?  There is a difference.  Filtering
 mail does not train SpamBayes.  Training mail does not filter it.

---------------------------------------------------------------------------
Peter Beckman                                                  Internet Guy
beckman at purplecow.com                             http://www.purplecow.com/
---------------------------------------------------------------------------



More information about the Spambayes mailing list