[SA-exim] Autolearn = No

[Magnus Holmgren]

On onsdagen den 13 februari 2008, Magnus Holmgren wrote:
> On onsdagen den 13 februari 2008, Jonathan GF wrote:
> > Hi group,
> >
> > the SA headers show i don't have enabled auto-learning but in my
> > spamassassin "local.cf" i have it (at least in term of bayes).
> >
> > Can someone show me he procedure to enable autolearning, and maybe pros
> > and cons ?
>
> autolearn=no means that autolearning was enabled but the score lay between
> the autolearn thresholds, or, if it was over the spam autolearn threshold,
> either the header score or the body score was below the hard-coded
> threshold of 3.0. Note that when SA decides whether to autolearn, the score
> is calculated using one of the non-bayes scoresets, i.e. it differs from
> the one that is reported in the X-Spam-Status field.

Oh, it could also mean that you haven't learned enough spam and/or ham for 
autolearning to kick in (bayes_min_ham_num and bayes_min_spam_num, which both 
default to 200).

-- 
Magnus Holmgren        holmgren at lysator.liu.se
                       (No Cc of list mail needed, thanks)

  "Exim is better at being younger, whereas sendmail is better for 
   Scrabble (50 point bonus for clearing your rack)" -- Dave Evans
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part.
Url : http://lists.merlins.org/archives/sa-exim/attachments/20080213/cad1ccc3/attachment.pgp