Train Bayes with sa-learn --ham using a sizable representative sample of the shipping company's "known good" email.
> -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] On > Behalf Of Håkon Nilsen (Exinet AS) > Sent: Friday, October 17, 2003 6:21 AM > To: [EMAIL PROTECTED]; > [EMAIL PROTECTED] > Subject: [SAtalk] LOTS of mail being tagged wrong > > Hi, > > This is a cross-message for MIMEdefang and SpamAssassin mailing lists. > > System: RH 7.3, Sendmail 8.12.10, SA 2.60, MIMEdefang 2.37 > > I just added new domain relaying through my mailserver. The > users are ship brokers, and receive a _lot_ of mail from > around the world. The problem is that lots and lots of the > mails are tagged as spam. > > 90% of the mails tagged as SPAM isn't spam. And that's not > very good. Here's the results of a mail: > > 5.4 BAYES_99 BODY: Bayesian spam probability > is 99 to 100% ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. SourceForge.net hosts over 70,000 Open Source Projects. See the people who have HELPED US provide better services: Click here: http://sourceforge.net/supporters.php _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
