low scoring spam

[Julian Field]

To a large extent, the Bayes database in SpamAssassin will teach itself.
You don't actually need to do anything, except when it gets it wrong. Then
you need to feed messages to "sa-learn". I'll leave others to explain how
to use sa-learn, but there have been many discussions on this here before.

But left to its own devices, SpamAssassin uses all its other rules to work
out what is definitely spam and definitely non-spam, and feeds those
definite messages back into the database learning code.

So it trains itself. Neat huh?

At 20:37 03/03/2004, you wrote:
>Thanks Steve,
>
>I've got MailScanner v 4.26.8-1.
>
>Bayes is running automatically. To say that I am *using* it may be
>overstating. I don't know how to train Bayes. I've read that you are
>supposed to feed it using sa-learn, but it is not clear to me exactly
>how that is done. I've been trying to figure it out from the FAQ. It
>doesn't help that I don't know my way around Linux very well.
>
>I changed servers recently and the old Bayes database was copied to the
>new server.
>
>I don't think DNSBLs are used, but I am not sure.
>
>Cathy Cramer
>
>
>
>
>Steve Thomas wrote:
>
>>On Wed, Mar 03, 2004 at 11:52:58AM -0700, Cathy Cramer is rumored to have
>>said:
>>
>>>I am having a real problem with random word spam receiving a spam score
>>>zero or very low, less that 3. Lots of this type of spam is getting
>>>through, while many legitimate messages get scores over 4. Some of my
>>>users are getting a hundred or more spam messages per day, about 90% of
>>>their total incoming mail. Are other people having problems with this?
>>>Any suggestions?
>>
>>
>>Are you using bayes and the DNSBLs?
>>
>>
>>--
>>"Logic is in the eye of the logician."
>>- Gloria Steinem

--
Julian Field
www.MailScanner.info
Professional Support Services at www.MailScanner.biz
MailScanner thanks transtec Computers for their support
PGP footprint: EE81 D763 3DB0 0BFD E1DC 7222 11F6 5947 1415 B654