Bayes and spam increase?

[Jeff A. Earickson]

This sounds like the "bayes poisoning" issue that has been discussed
numerous times on this list.  I've kept the following in my
spam.assassin.prefs.conf file:

score BAYES_00 0 0 -0.05 -0.05
score BAYES_01 0 0 -0.04 -0.04
score BAYES_10 0 0 -0.03 -0.03
score BAYES_20 0 0 -0.02 -0.02
score BAYES_30 0 0 -0.01 -0.01

I don't trust Bayes enough to let it substantially lower a score --
only to increase a score.

Jeff Earickson
Colby College

On Mon, 7 Feb 2005, Magda Hewryk wrote:

> Date: Mon, 7 Feb 2005 13:22:41 -0500
> From: Magda Hewryk <MHewryk at SYMCOR.COM>
> Reply-To: MailScanner mailing list <MAILSCANNER at JISCMAIL.AC.UK>
> To: MAILSCANNER at JISCMAIL.AC.UK
> Subject: Re: Bayes and spam increase?
>
> Yes, I've got a lot  untagged spam email on the weekend.  I found BAYES_00
> -2.60 attached to all of them.
>
>
> Thanks,
>
> Magda
>
>
>
>             Matt Kettler
>             <mkettler at EVI-INC
>             .COM>                                                      To
>             Sent by:                  MAILSCANNER at JISCMAIL.AC.UK
>             MailScanner                                                cc
>             mailing list
>             <MAILSCANNER at JISC                                     Subject
>             MAIL.AC.UK>               Re: Bayes and spam increase?
>
>
>             02/07/2005 11:23
>             AM
>
>
>             Please respond to
>                MailScanner
>               mailing list
>             <MAILSCANNER at JISC
>                MAIL.AC.UK>
>
>
>
>
>
>
> At 10:43 AM 2/7/2005, Fractal IT Dept. wrote:
>> Hi everyone!
>>
>> We've noticed an increase in the number of spam sneaking through with
>> scores "just under" our threshold. After looking through the headers for
>> these messages, I've noticed that bayes seems to have "no opinion" on the
>> majority of these (ie. no bayes entry). Am I missing something? I thought
>> bayes would score every message?
>
> That's not entirely true, especially for the 2.6 series.. in 2.6x or 2.5x,
> In those any "no matches" or other 50/50 chance does not get a BAYES_ rule
> match.
>
> Can you tell us what version of SpamAssassin you are using?
>
> ------------------------ MailScanner list ------------------------
> To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
> 'leave mailscanner' in the body of the email.
> Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
> the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).
>
> Support MailScanner development - buy the book off the website!
>
> ------------------------ MailScanner list ------------------------
> To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
> 'leave mailscanner' in the body of the email.
> Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
> the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).
>
> Support MailScanner development - buy the book off the website!
>

------------------------ MailScanner list ------------------------
To unsubscribe, email jiscmail at jiscmail.ac.uk with the words:
'leave mailscanner' in the body of the email.
Before posting, read the MAQ (http://www.mailscanner.biz/maq/) and
the archives (http://www.jiscmail.ac.uk/lists/mailscanner.html).

Support MailScanner development - buy the book off the website!