I think you need to reread Paul's thesis. Rather than using fifteen
keywords, his filter calculates the "spam value" of the fifteen *most
significant* words found in the message. IOW, it look at *all* the
strings in the message - header and body - assigns values to each and
every string and then sums the values of the fifteen most significant
words.

I think the single most significant aspect of his research is the 0
false positives result. That alone makes it worth pursuing.

Paul Schmehl (paulsutdallas.edu)
Project Coordinator
University of Texas at Dallas
http://www.utdallas.edu/~pauls/
AVIEN Founding Member
-----Original Message-----
From: Greg Hackney [mailto:hackneyswbell.net]
Sent: Sunday, September 01, 2002 2:33 PM
To: postfix-userspostfix.org
Subject: Re: Using blacklists and RBL's with Postfix

To be impressed I'd have to see it tested on a system with hundreds of
thousands users
with a dozen different spoken languages, and run through a couple
million spam messages.
I thought SpamAssassin (still in it's infancy) was already light years
ahead of Paul.

To me all Paul is saying basically is that gee, spam blocking ought to
use some intelligent
algorithm for content filtering, and Bayes looks promising. Nothing
newsworthy to me.
 
-
To unsubscribe, send mail to majordomopostfix.org with content
(not subject): unsubscribe postfix-users

• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]