Internet Chicago Staff wrote: > So how would I "appropriately" know what e-mail has been misclassified as > SPAM when its not or HAM when it is not ? > > Your not suggesting that the Bayesian test is perfect are you ?
Initial training of ASSP's corpus is what the Bayesian test-mode is for. I'm not sure what you mean by "perfect", as the Bayesian test-mode simply lets all Bayesian spam through. Refining the corpus is what the email interface is for, and its generally the recipient that drives this functionality. You don't have to manage it as an administrator, your users (and you too if your received spam) do on an individual basis. If you really want to perform this as an administrator, then use the sendAllSpam address to capture copies of your spam. You should be able to easily filter Bayesian spams from other spam-types based on the message headers. You can then forward the Bayesian spam back to ASSP for corrections. Your end-users (the recipients) should drive all false-negative spam (spam classified as ham) back to ASSP for corrections. This is by design, and is how your users become self-sufficient - taking administrative overhead away from you. The three principle email addresses you should be using for this are EmailSpam, EmailHam, and EmailWhitelistAdd - but you should review and consider using all the addresses provided for maximum ease of use. ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Assp-user mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/assp-user
