Gordon
Weakliem: I've just installed Mark Hammond's Bayesian Spam
Filter plugin for Outlook, part of the Python
spambayes
project.
Now that looks exciting and versatile! It looks like it can be
used with procmail, so I can use it too, not just with my e-mail
but also with my weblog.
It is interesting to note that I have received zero
comment spam since I installed SpamAssassin. It wasn't filtered out
- it never came.
An initial update - I also installed this on a co-worker's machine, she receives around 50 spams daily, and more on the weekends. We trained it with around 800 known good messages and 6 spams. In the last hour, she's received another 6 spams, 2 which went straight to the spam folder, and 4 which were marked probable. I looked at the analysis of one of the probables; the word 'click' ranked about .84 spam probability. After I dragged the 4 probables to the spam folder, 'click' now had a spam ranking of ~.86. I know I shouldn't rave on like this, but this thing is _working_!
We were also impressed with the SpamBayes project, so we decided to make a product out of it. Basically, it is designed for consumers that don't want to install Python or do other things that are not usual for consumers.
Sorry for the clear commercial plug, but we really want to get the SpamBayes message out. Our site is www.inboxer.com. We have a free trial and we would love feedback.