Øystein Gisnås wrote: > This mail made it 4.1 on my score meter (default settings v2.54, 5.0+ > triggers): > > Subject: AMBlEN: $100 and VlAGRA: 39.95 > From: Online Drug Sale <[EMAIL PROTECTED]> > "On sale for today only: VlAGRA and AMBlEN > > It scored on BAYES_90=3.002, CLICK_BELOW=0.0, RCVD_IN_SBL=1.101. I > feel that this particular mail should have been punished for > mentioning "online drug sale", V1AGRA, AMB1EN and pills so many > times. What about including the word "V1AGRA" in a viagra-rule? Or > even better, make a separate, more heavily weighted rule for it. > "V1AGRA" is more seldom used than "VIAGRA", except for in spam. >
Is your Bayes trained and working? I think my SA would have scored that like 40+... Bayes anyways is the tool for that: you can't make a separate rule for each an all "nasty" words there;) With all default rule weights, several black lists activated, well trained with ham and spam, most of my spam is in range 20-50 points. If it gets under 5 it most propably is a ham. ------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
