Mike Burger writes: > Agreed...it's not likely that that particular message is going to pass > through the list, again, in the same form, so the bayes database wouldn't > really mean squat at that point, insofar as that message goes.
Yes -- but when you learn a message as ham, it learns *all aspects* of the message -- including the message sender, the network it came from, boilerplate text in the body etc. Plus if the sender has a habit of talking about "erectile dysfunction", you *want* that phrase to have a less spammy overall meaning than the default SpamAssassin ruleset would give it -- so counterbalancing with bayes is a good plan. So it would have a good effect. The bayesish classifier is good at this stuff! --j. ------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk