On Sunday, March 09, 2003 at 07:55:41GMT -0500 (which was 7:55 AM where I live) Foster, Graham wrote and made these points on the subject of "Bayesian spam filtering": FG> I've just discovered that Mozilla has included THE most powerful spam FG> filtering system available - Bayesian filtering as part of its e-mail FG> client. It comes as standard. Nice to see this technology entering the FG> mainstream. Pity that TB! plug-architecture will not easily support Bayesian FG> filtering, as you need a dynamically maintained "bad" and "Good" lists. Once FG> each are of a sensible size you start the Bayesain filter, and it FG> characterises the arriving email as Spam - or not. (all Stats based). I FG> don't see how this approach will fit in with the intened keyword based FG> Anti-Spam filters in TB! FG> Anyway.. if you are interested.. see FG> http://www.mozilla.org/mailnews/spam.html FG> Graham
I prefer using SpamPal for doing the spam filtering as it can query spamcop and other services, has black and white lists, and has plugins to handle other filtering methods (such as Bayesian). -- Best regards, David Adolescence and snow are the only problems that disappear if you ignore them long enough. Using The Bat! v1.62i on Windows 98 4.10 Build 2222 A ________________________________________________________ Current version is 1.61 | "Using TBTECH" information: http://www.silverstones.com/thebat/TBUDLInfo.html
