Hello

I tested it some day  1.9.5.0 (0.0.10) with  rebuildspamdb 2.9.5.0.  I noticed 
an increase of spam passing (from none/rare to 6-8 per day)
and the reason is always due to the negative score (I suppose due to 
spamdberror) .

X-Assp-Score: 43 (Bayesian Probability: 0.9825)
X-Assp-Score: -9 ([scoring:-9] Bayesian Reports Probability: 0.0000)

probably is better to use spamdberror only to assign positive  X-Assp-Score 
scores
(avoiding negative scores which are making passing more spam than usual) .

Graziano

> There is new feature in 1.9.5.0 (0.0.10).
> (Needs rebuildspamdb 2.9.5.0)
>
> The spam/ham reports will be used to build a second spamdb called
> spamdberror.
> A second Bayesian analysis will be done based on this second spamdb.
> baysValencePB_error is weighted with the second probabilty if>=
> spamprob or<  1-spamprob.
>
>
> ------------------------------------------------------------------------------
> RSA(R) Conference 2012
> Save $700 by Nov 18
> Register now
> http://p.sf.net/sfu/rsa-sfdev2dev1
> _______________________________________________
> Assp-test mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/assp-test
>


------------------------------------------------------------------------------
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
http://p.sf.net/sfu/rsa-sfdev2dev1
_______________________________________________
Assp-test mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-test

Reply via email to