On Thu, 12 Jul 2007, Diptanjan wrote:

> I have set the German rules and its updating from a chennel
> file... however, mails are not regarded as spam because the bayes
> check rates it as non-spam, giving a high (-)tive score

Those two facts are unrelated.

Bayes scores based on what is in its token database. The other rules
don't affect the score it generates.

> Anyone have any suggession? Plz help me.

At first glance your bayes appears to be mistrained, however it is
possible that the random text in some stock spams will happen to look
hammy to your bayes. Does the non-stock-scam text in that message look
like your regular legitimate traffic?

(1) How often does this happen?

(2) How big is your bayes database? (use "sa-learn --dump magic")

(3) How are you training your bayes database? Autolearn? Or manual?

(4) If manual, do you keep your corpus around after you've trained it?

(5) If manual, do you let nontechnical users train it without review?

Finally, have you installed any of the SARE rules from
www.rulesemporium.com? While not "fixing" a low bayes score, they may
counteract it when bayes gets fooled.

John Hardin KA7OHZ http://www.impsec.org/~jhardin/
jhardin@impsec.org FALaholic #11174 pgpk -a jhardin@impsec.org
key: 0xB8732E79 -- 2D8C 34F4 6411 F507 136C AF76 D822 E6E6 B873 2E79
Ignorance doesn't make stuff not exist. -- Bucky Katt
12 days until The 38th anniversary of Apollo 11 landing on the Moon