Bayes Question

[Kevin Hanser <Ke...@mica.net>]

Lately I've noticed that there seems to be more spam than usual creeping
through our SpamAssassin installation.  I looked @ the message headers,
and I've noticed that on most of the ones that are getting through, I'm
seeing this:
 
BAYES_00 -4.90
 
What does that mean?  Do I have an improperly trained Bayes Database
somehow?  Is there a way that I can tell SA to reset this one Bayes
rule?
 
Thx!
 
k

Re: Bayes Question

[Matt Kettler <mk...@evi-inc.com>]

At 12:11 PM 2/20/2004, Kevin Hanser wrote:
>Lately I've noticed that there seems to be more spam than usual creeping 
>through our SpamAssassin installation.  I looked @ the message headers, 
>and I've noticed that on most of the ones that are getting through, I'm 
>seeing this:
>
>BAYES_00 -4.90
>
>What does that mean?  Do I have an improperly trained Bayes Database 
>somehow?

Generally, yes... You definitely should not be hitting BAYES_00 on spam. 
Looking at my most recent 200 spams, the lowest any of them scored is BAYES_44.


>Is there a way that I can tell SA to reset this one Bayes rule?

No.. Not in any practical way at least. Bayes isn't directly rule based...

Bayes generates a spam-probability percentage from 0 to 99. BAYES_00 means 
that the bayes calculation resulted in less than 1% probability of 
spam.  The way to reset the bayes calculation is to reset the entire bayes 
database.

You could change the score of BAYES_00, but that's covering up the effects, 
and not solving the problem.. you've got a poorly trained bayes database.