Hello, K.. You wrote 12.10.2004 @ 11:36
KS> Thanks for the mail. DO you mean that even my legitimate mails are caught KS> as spam because I have not taught bayes more ham against spam. (the ratio KS> of my ham:spam training is 2:100) Yes. It works very like a human. You "explain" to the bayesian-based filter "what is" spam and ham by "training" it. It you train 100 spams and 2 hams, the filter very well "knows" what is spam, but is practically useless in the case of hams. And it will work accordingly with it's "knowledge" - i.e. it will grade the most letters by the way it "knows" better - namely spam. -- Sincerely, Alexey. Using TB 3.0.1 on WinXP Pro SP2 (2600), spelling by ORFO2002 (CSAPI) ..with Kaspersky Antivirus Plugin (ver 3.5 Gold) & antispam filter BayesIt! 0.7.4 mailto:[EMAIL PROTECTED] ________________________________________________________ http://www.silverstones.com/thebat/TBUDLInfo.html