-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi,
I used to use spampal and popfile. I found that mail collection was very slow and sometimes hung both the bat and the proxy program. So, back when BayesIT was 5.3 or so, I decided to try it. I did not train it at all. As emails came in, I categorized them as junk or not. After three days or so, it was up to 98% accuracy. It still makes a mistake, but it is 1 per 1000 emails or less. In addition, and this is what is amazing, BayesIT has sped up email collection by at least 300%. The speed difference is noticeable and amazing. No Bayes based filter will ever be 100% perfect. I believe, however, that BayesIT is a very good implementation of the Bayes calculations. Since the 5.10 or so, I have added the discussion and newsletter groups to the whitelist, which has sped up mail collection even more. This list generates a hundred or more email a day, and those are just passed through BayesIT, and not counted in the totals. - -- Michael L. Wilson, MBA Ecclesiastic Philosopher Critic Teacher :einsteinyoyo: Rant about The Bat! :rant: [EMAIL PROTECTED] ====================================================== Using The Bat! 3.0.0.7 Windows XP 5.1 (2600-Service Pack 2) Ponder this: Fool me once, shame on you. Fool me twice, shame on me. ============================================= Spam Stats, last 24 hours (BayesIt! 0.6.01) Total Spam Emails: 69 Total Clean Emails: 15 BayesIT guessed right 99.971014% of the time My email is 82.142857% spam ============================================= morning day evening ============================== 17:50 -----BEGIN PGP SIGNATURE----- Version: PGP SDK 3.2.2 Comment: "" iQA/AwUBQT0HSG2u7N+Qh+ieEQInmwCfa52NHboFJYrJ0PRnPMh6CBV+UyEAoPE3 NQbo7IAfYY1ZK4GEeO99VaMg =rQUe -----END PGP SIGNATURE----- ________________________________________________ Current version is 3.00.00 | 'Using TBUDL' information: http://www.silverstones.com/thebat/TBUDLInfo.html
