Re: Strategy for collecting spam to feed Bayes?
Bill Cole <[email protected]>
| Newsgroups | gmane.mail.spam.spamassassin.general |
|---|---|
| Message-ID | <[email protected]> |
On 2025-01-13 at 03:12:25 UTC-0500 (Mon, 13 Jan 2025 02:12:25 -0600 (CST)) Dave Funk <[email protected]> is rumored to have said: > It's also possible for the messages to differ by things such as > network routing headers, better to feed it all to bayes and let it get > parsed/scored. That's an important fact. One of my most active traps frequently will get 3 identical-body spams via 3 different cracked-host routes in under 5 hours when one of my pet spammers (mostly offering office products in German) switches their content substantially. Feeding those straight to Bayes means that the ~100 additional copies for real users get rejected no matter which path they arrive by. -- Bill Cole [email protected] or [email protected] (AKA @[email protected] and many *@billmail.scconsult.com addresses) Not Currently Available For Hire