I have been wondering about how to improve spam classification for my clients and ended up with an idea that should be possible with amavis.
I have been using only spamassassin for spam anslysis and I *really* love the possibility I have to improve that due to its rules. Besides this, I not so sure about its bayes effectiveness. I usually deploy it using wide-level settings instead of user-level settings for bayes and I know this will reduce its classification power... But, the major drawback is classification feedback from users due to wide settings and many different flavors of e-mail clients, web interfaces etc. So, how about take advantage of dspam native classification method and keep spamassassin only with rules, like this: Amavis grabs email message and pass it thru SA rules to score it. If the score - according to your settings - is high, set it as spam and classify it as spam in dspam. If not, pass it thru dspam and classify it as legitimate. (User can set it as spam later, for instance). Using this approach, users will have the (nice) dspam interface to report spam/ham and user feedback about classification is pretty straightforward. Let me know what you people think of it. Regards, Neto. ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ AMaViS-user mailing list AMaViS-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/amavis-user AMaViS-FAQ:http://www.amavis.org/amavis-faq.php3 AMaViS-HowTos:http://www.amavis.org/howto/