Bayes: Worth it?

[Jan-Peter Koopmann]

    [ The following text is in the "iso-8859-1" character set. ]
    [ Your display is set for the "US-ASCII" character set.  ]
    [ Some characters may be displayed incorrectly. ]

Hi,

> Please excuse my bayes ignorance, but don't you have to train
> it in some way to know what is really spam and what is really ham?
> How do you train a system like this when it's acting as a
> gateway?  I always assumed that there had to be user
> intervention in the training process...


Yes and no. As Julian pointed out, SA can auto-learn. You can also teach it via Exchange. One way would be to use public folders SPAM/HAM into which users drag their mails. You can then fetch the mails out of there and feed it back to SA via sa-learn. We currently archive all incoming mail for three days on the proxy via Mailscanner, fetch mails from Exchange public folders, lookup the original mail using the Message-ID and then feed the original mail to SA, DCC, Razor, Spamcop.

Regards,
  JP

------------------------ 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!