Hello, A> I suspect that since the Bayesian filter learns, you may not want it A> learning stuff that you have other means of detecting. All I think you A> will do is make it try to duplicate the work done elsewhere, and it A> will be filled with words it does not really need or want to know A> about.
That's a good reflexion... I'm not good enough with Bayesian filter to know what's the good solution... But not marking SPAM as Junk will also fill the filter with bad data ? Am I wrong ? -- Cedric - mailto:[EMAIL PROTECTED] http://www.haute-vitesse.org (DH/DSS)PGP-key Server ID: 0xBDD6E604 ________________________________________________ Current version is 2.01 | "Using TBUDL" information: http://www.silverstones.com/thebat/TBUDLInfo.html