I'd say 98.4% accuracy is pretty darned good, Robin. Hold off a little, take the missed spam and send it through sa-learn. I don't know where you got the corpus of spam and ham from, but as you send current traffic through sa-learn, rather than a slightly stale corpus, or a corpus from another user, it'll get more and more accurate. I wouldn't go messing with things that are 98.4% accurate from the get-go. It'll only get better. You'll see the Bayesian classifier start giving higher confidences on hammishness and spammishness.
> > -----Original Message----- > From: Robin Whittle [mailto:[EMAIL PROTECTED] > Sent: Monday, June 09, 2003 3:27 AM > To: [EMAIL PROTECTED] > [ ] > > I am about to raise the scores of Bayes results which are > higher than 50%, because I think that raising these scores > will reduce the number of false negatives. I will probably > find out the answer to this question by doing this. I spent > some time looking through the documentation and didn't find an answer. > > > Thanks for SpamAssassin! > > - Robin ------------------------------------------------------- This SF.net email is sponsored by: Etnus, makers of TotalView, The best thread debugger on the planet. Designed with thread debugging features you've never dreamed of, try TotalView 6 free at www.etnus.com. _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk