Hello Barry, > I've now transcribed 83 different regex rules over to Bayesit. I extracted > them from my copy of MailWasher. > This seems to have made a dramatic improvement in spam catching. I've now > reduced the spam failures from 100% (really!) to below 54% in two days.
this seems to be a great deal of work just to filter spam. I set up K9 and had it at 95% accuracy in about a week's training. It has improved since. Furthermore: I never got BayesIt to work. The fact that there doesn't seem to be documentation doesn't help. Even when the German magazine c't (highly respected) tested several mail programs and their spam filtering capabilities they couln't get it to work. In other tests that I've read BayesIt seemed to always be worse than other Bayes-based competitors. The upshot of this: "No BayesIt forrrr me!!11" Roman -- We are on the verge: Today our program proved Fermat's next-to-last theorem. ________________________________________________ Current version is 3.0.1.33 | 'Using TBUDL' information: http://www.silverstones.com/thebat/TBUDLInfo.html