I have a script that I use with Exchange/Outlook for Bayes training, but it's not simple. You can't just forward a message back to the SA box, since Outlook deletes most of the original headers. You have to "cut-n-paste" the whole email into a new email and send THAT to the SA box. There the script unencapsulates the email and feeds it to sa-learn.
I don't consider my script to be distributable, and it's only part of a larger scheme involving DNS, sendmail redirection, and other variables, but at least this might give you some ideas to play with. Pierre Thomson BIC -----Original Message----- From: Jason Gauthier [mailto:[EMAIL PROTECTED] Sent: Thursday, January 06, 2005 8:44 AM To: Jeff Koch; users@spamassassin.apache.org Subject: RE: Bayes FP/FN Training Procedures Neat! I was just thinking about how to do that myself. But, I use exchange, so I'm not sure how to do it yet. > -----Original Message----- > From: Jeff Koch [mailto:[EMAIL PROTECTED] > Sent: Thursday, January 06, 2005 8:42 AM > To: users@spamassassin.apache.org > Subject: Bayes FP/FN Training Procedures > > > Has anyone come up with a script or method that would allow > users to forward their false positive and false negative > emails back to an address on the mailserver where they can be > used to train the Bayes database. I understand that Bayes > needs the email in its original format so the script has to > strip off the forwarding enclosure. > > Thanks in advance. > > > > > Jeff Koch > > >
