On 2010/04/29 8:25 AM, Frank Bures wrote:
I've been running spamassassin for years. I am using auto-learn with very conservative thresholds. However, after several years of usage my spam database is about three time larger than my ham database and I am starting to see false positives.Is there a way how to "shrink" the spam database?
I use autolearn as well, and have seen the same lean toward spam since an upgrade to 3.3.x. If, like mine, your messages pass through the system to another then there is no way to "unlearn" some of the spam messages. Lately, I've been clearing the Bayes database and toying with the thresholds a bit to even things out. Still haven't got it just right, but these things take time.
-- /Jason
smime.p7s
Description: S/MIME Cryptographic Signature