Hi I have had the following issue with autolearning enabled:
I turned autolearning on with the default settings, but what started to happen was that the rate of mails being autolearned as ham far outstripped that of mails being learned as spam. The result appears to have been once a critical point was reached, large amounts of spam started to be classified as ham. Is it simply a matter of playing around with the autolearn thresholds until I find a good balance? I would dearly like to utilise autolearning as I am running SA in a web hosting environment and so have numerous servers hosting even more numerous domains all talking to the same server via spamc/spamd and hence utilising the same bayesian database. If I can get autolearning working then this seems to be the right way forward for me. Thanks Deon.
