On Sat, 07 Feb 2009, Ricardo Kleemann wrote:
> I have SA working very well for me, but there are still a few cases of spam
> that are very persistent, I still get a considerable amount of spam that SA
> doesn't catch.
>
> However, what is annoying is that no matter how much I feed through
> sa-learn, the SA score doesn't change much.
>
> I created a folder sa-learn where every day I'll put in the spam that is
> not caught, and once a day I have a cronjob that automatically feeds the
> messages to sa-learn, and every day I get the confirmation that X messages
> were learned.
>
> So anyway, what does it take for sa-learn to actually start changing the
> score so that eventually I can get the correct classification?
Do you run sa-learn as the same user who is performing the SA checks on
incoming mail? Does your Bayes database contain at least 200 (of each) spam
and ham? What is the output of:
% sa-learn --dump magic
Also read: http://wiki.apache.org/spamassassin/BayesNotWorking.
--
Sahil Tandon <sa...@tandon.net>
