although I have few thousands of messages in my spam and notspam
directories, sometimes I see messages coming with very high
X-Assp-Spam-Prob and zero X-Assp-Bayes-Confidence. It happens with
messages with an attachment and no text in body.

I have set Bayesian Confidence Level so that this messages are just
marked and not rejected but my users get nervous about receiving spam
marked messages for no reason.

It would be nice to have a minimal bayesian confidence threshold
below which beyesian result is ignored.
I'd like to do this:
confidence < 0.001   messageok
0.001 < confidence < 0.05   mark as spam
confidence > 0.05   reject message

I think this vould be very useful for new installations

I'm not a byesian guru so i don't know if I'm missing something...

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Assp-user mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-user

Reply via email to