I’m now using a two-tiered approach with SpamAssassin at the server and SpamBayes on the client. SpamAssassin has been great but it’s a constant battle where the spammers periodically figure out how to slip more under its nose by tweaking the content patterns. So I put SpamBayes’ plug-in into Outlook 2k and it’s been doing an excellent job in the first few days of nabbing the runaways.
Check out http://spambayes.sourceforge.net/ and starting using it. Training on my message base seemed to hang after awhile, but I restarted Outlook and it’s been fine, learning from what I do with messages it thinks may be spam. It seems to err more on the side of false positives, but hasn’t missed any real spam yet.
No related posts.