I would set up 2 aliases: [EMAIL PROTECTED] --> |/usr/bin/sa-learn --spam --single [EMAIL PROTECTED] --> |/usr/bin/sa-learn --ham --single
The users would forward the false-positive to spam and the false negative to nospam.
My question is, since the repported mail will be a forwarded mail, will spamassassin also learn the email address of my users and the subject (fwd: original-subject) and the rest of the headers? I wouldn't want the users to see all the mails coming from them being marked as spam just because they repported lots of false negative...
No, it won't learn their addresses, but it WILL learn that all forwarded messages are spam :)
It will also learn things about their mailclient, your choice of message ID's, etc, etc.
Forwarding this simplistic way DOES NOT WORK for bayesian learning.. it's in the FAQ flat out that you can't do this.
If there a simple way to make SA learn only the body of the repported mail?
No, because SA heavily tokenizes the headers.. it doesn't just examine the body when doing scoring, thus cannot examine just the body when doing learning.
does sa-learn strip the spam repports from a mail when submitted?
yes.
------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
