On Tue, 27 Jul 2004 18:23:42 -0600 (CST), Ryan Thompson <[EMAIL PROTECTED]> wrote: ... snip ... > 3) Time. By asking your users to spend time hand-classifying and > training their filter, how does this save them from having to look at > spam? Contrarily, if you spend a few hours a week training a > site-wide filter, you can easily save *hundreds* of person hours for > your users. As your userbase increases in size, this becomes a *huge* > win. That's what I'd sell as one awesome value-added benefit. ... snip ... > So, given the above, I endorse site-wide Bayes for almost all large > installations. The practical advantage of per-user bayes is highly > debatable, and, even if there *is* a practical advantage, it is probably > small, and not worth the time required.
AWESOME info, Ryan. I really appreciate the time & insight. Just curious - for site-wide Bayesian training, then: - Are you telling your users to send all spam to one central address, that you then use to teach your site-wide Bayes as spam? - What are you feeding Bayes as ham? Or is that part of your Squirrelmail setup - so that all that's not spam is ham, when done? - (OR) - What are you training it with, if not emails from your clients/users? Thanks so much for your help & ideas. - Miles
