Date: Wed, 12 Jan 2005 10:52:23 -0500 From: Louis LeBlanc <FreeBSD@keyslapper.org> To: freebsd-questions@freebsd.org Subject: Re: SA, Bayes, and You (erm, Me) Message-ID: <20050112155222.GB85890@keyslapper.org> In-Reply-To: <5346C474-648E-11D9-AD37-000D9333E43C@secure-computing.net> References: <200501120509.j0C59Nu7084400@mail.cs.ait.ac.th> <5346C474-648E-11D9-AD37-000D9333E43C@secure-computing.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On 01/12/05 05:37 AM, Eric F Crist sat at the `puter and typed:
> On Jan 11, 2005, at 11:09 PM, Olivier Nicole wrote:
>
> >> However, when I read the
> >> headers of any messages, I notice that they ALL say 'autolearn=no',
> >> even though I've enabled autolearn in local.cf (bayes_auto_learn 1).
> >
> > There is something with the threshold for auto learning.
> >
> > olivier
>
> I've got the threshold turned down to 6.0. Still nothing.
There's still more to it than that. Mail::SpamAssassin::Conf(3) has
the following:
bayes_auto_learn_threshold_spam n.nn (default: 12.0)
The score threshold above which a mail has to score, to be fed into
SpamAssassin's learning systems automatically as a spam message.
Note: SpamAssassin requires at least 3 points from the header, and
3 points from the body to auto-learn as spam. Therefore, the mini-
mum working value for this option is 6.
So, a message can score 25 points off the body and 2 off the headers,
and it will not be autolearned.
Lou
--
Louis LeBlanc FreeBSD@keyslapper.org
Fully Funded Hobbyist, KeySlapper Extrordinaire :)
http://www.keyslapper.org ԿԬ
Abbott's Admonitions:
(1) If you have to ask, you're not entitled to know.
(2) If you don't like the answer, you shouldn't have asked the question.
-- Charles Abbot, dean, University of Virginia
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050112155222.GB85890>