For people looking for a personal antispam solution, perhaps SpamBayes http://spambayes.sourceforge.net/ might be an idea. Bayesian filtering is great because it adapts to the user. Ideally, to get it to work decently out of the box, one should save mail for a while first - sort all spam into a folder and keep the good mail in another folder (could be the inbox.) Then, one installs SpamBayes and feeds it the good stuff and the bad stuff as something to build on and it will have a good baseline of what one likes and what one doesn't. You'll then continue marking things as either spam or ham (ham being the good mail) and the filter keeps learning and updating itself. I hear accuracy percentages in the high 90's mentioned, though I haven't had the opportunity to use it much (I get extremely little spam to my work account, and for personal stuff I use Gmail which filters nicely.) The easiest way to use SpamBayes is if one uses Outlook, there is a nice plugin for it. Can be used with other programs, but it's not as nicely integrated then. Article from 2003 about it. Since I haven't used it much I have no idea how well it works these days, but it stands to reason the underlying idea hasn't changed.