It's the only version that didn't cause a ton of false positives. For instance, assume someone uses their first name as their username: [EMAIL PROTECTED] They'll get legitimate mails like "please update this spreadsheet bob ASAP", or "bob, let us know if you can goto lunch", etc.
I've never seen a legitimate email like that. No one I know goes to the trouble of typing my name in the subject line. They might in the first line of the body if there's a lot of recipients, but that's a different issue.
Yet I get tons of spam with "Kenneth" as the first word (what follows varies, might be comma-space, might be my last name).
Do recent entries in the corpus show this kind of FP? Have you looked for trends in the corpus that show legit users modifying their behavior to avoid looking like spammers?
------------------------------------------------------- This SF.net email is sponsored by: IBM Linux Tutorials. Become an expert in LINUX or just sharpen your skills. Sign up for IBM's Free Linux Tutorials. Learn everything from the bash shell to sys admin. Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
