Bayes learn problem

Sam Przyswa samp at arial-concept.com
Mon Dec 18 17:17:16 CET 2006


Hi,

I have a problem with Spamassassin and Bayes the command "sa-learn -p 
/etc/MailScanner/spam.assassin.pref --dump magic" give me:

0.000          0          3          0  non-token data: bayes db version
0.000          0      37411          0  non-token data: nspam
0.000          0      82399          0  non-token data: nham
0.000          0     142717          0  non-token data: ntokens
0.000          0 1163567938          0  non-token data: oldest atime
0.000          0 1166315229          0  non-token data: newest atime
0.000          0 1163774133          0  non-token data: last journal 
sync atime
0.000          0 1163774206          0  non-token data: last expiry atime
0.000          0     234789          0  non-token data: last expire 
atime delta
0.000          0          0          0  non-token data: last expire 
reduction count

Then after found more than 180 spam mails in my inbox with Thunderbird I 
put these mail in a learn directory and run the command "sa-learn -p 
/etc/MailScanner/spam.assassin.pref --spam <learn directory>" I got:

Learned tokens from 106 message(s) (188 message(s) examined)

And then the command "sa-learn -p /etc/MailScanner/spam.assassin.pref 
--dump magic" give me:

0.000          0          3          0  non-token data: bayes db version
0.000          0      37411          0  non-token data: nspam
0.000          0      82401          0  non-token data: nham
0.000          0     142717          0  non-token data: ntokens
0.000          0 1163567938          0  non-token data: oldest atime
0.000          0 1166315229          0  non-token data: newest atime
0.000          0 1163774133          0  non-token data: last journal 
sync atime
0.000          0 1163774206          0  non-token data: last expiry atime
0.000          0     234789          0  non-token data: last expire 
atime delta
0.000          0          0          0  non-token data: last expire 
reduction count

No new spam or tokens added ?

Today SA found 1511 spam mails but about 180 are not deterted.

When I test a learned spam message with the command "spamassassin -p 
/etc/MailScanner/spam.assassin.pref -D < learned_spam_message' I got :

...
X-Spam-Checker-Version: SpamAssassin 3.1.0 (2005-09-13) on
        neuilly.neuilly.geismar.com
X-Spam-Level: *
X-Spam-Status: No, score=1.2 required=5.0 tests=AWL,BAYES_50
        autolearn=unavailable version=3.1.0
...

 From the header checked mail, the score is only 1.2 for a learned spam 
mail !?

What's wrong ?

Sam.


-- 
Ce message a été vérifié par MailScanner
pour des virus ou des polluriels et rien de
suspect n'a été trouvé.



More information about the MailScanner mailing list