Quoting Matt Kettler <[EMAIL PROTECTED]>:
At 10:00 AM 7/14/2004, Jim Maul wrote:Its not 1000 of each email, its 200 (by default). If autolearning is enabled
(also default) you should see bayes kick in pretty quickly actually if your
spams are scored high enough and your hams are scored low enough. You could
train manually but that requires more work...its really up to you. I'd let it
autolearn for a bit and see how long it actually takes to reach 200..i dont
think it would take that long.
Ugh.. IMO, and in past opinons expresed by the developers, this is bad advice.
I have never manually learned a message in my system and its working quite well.
Above average i'd say. Never had a false positive and the only false negatives
i have had were before bayes was being used. Perhaps it doesnt work for some
people, but it seems to work really well this way for low volume servers (which
i have).
I'd agree that you can use autolearning to pick up some of the 200/200 messages, but I'd do at least *some* hand training.
Of course. I didnt say DONT do any manual training, i simply suggested that it
isnt necessary and that ultimately its "up to you" whether or not you do so.
The autolearner isn't perfect, and an autolearn-only bayes database has a noticable chance of ending up poison. It doesn't happen every time, but there's a distinct chance of it.
Anything can happen, and of course it doesnt happen every time. Again, i'd say
about 95% of the time my system is triggering bayes_99 or bayes_00. The so-so
rules (middle bayes where it isnt really sure) hardly ever get used and my
spams are usually above 15 while the hams are usually below -4. I'd say these
are pretty good results.
Giving it a small hand-trained head start helps prevent the autolearner from going awry due to the "never autolearn something that would strongly contradict existing bayes learning" rule.
To each his own.
Jim
