All, I have been seeing spam with a large number of random words, formated as sentences complete with punctuation. These are commonly getting by the SA Bayes filters, and are occasionally getting past the other checks to land in my mailbox.
I am curious on two fronts: First, the ones that are getting though, have the spammers found a reasonably effective way to combat Bayesian filters? Was Paul Graham and those who went before him wrong when they predicted the response to Bayesian filter was a short, compact e-mail with little content? I would if this could also be a way to combat checksome checks such as DCC and Rayzor - if the words are unique to the e-mail and no duplicated in the google they send out. Secondly, do the e-mails that get properly flagged as spam, and that are above the auto-learn threshold going into the Bayes database and having a negative impact on its ability to find and identify spam? I have been holding off adding these types of spam e-mail that has got through the filters into my SPAM mailbox (which I use sa-learn) on in case this may be the case. Interesting see if anyone has any insight. -- Kind Regards, David A. Flanigan [EMAIL PROTECTED] http://www.flanigan.net ------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
