You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@spamassassin.apache.org by jm...@apache.org on 2004/07/12 07:53:08 UTC

svn commit: rev 22834 - spamassassin/site

Author: jm
Date: Sun Jul 11 22:53:07 2004
New Revision: 22834

Modified:
   spamassassin/site/main.wmk
Log:
wow, forgot this feature for a while

Modified: spamassassin/site/main.wmk
==============================================================================
--- spamassassin/site/main.wmk	(original)
+++ spamassassin/site/main.wmk	Sun Jul 11 22:53:07 2004
@@ -135,6 +135,13 @@
     as <a href=http://www.mail-abuse.org>mail-abuse.org</a>, <a
     href=http://www.ordb.org>ordb.org</a> or others.
 
+  - **learning classifier**: SpamAssassin uses a Bayesian-like form of
+    probability-analysis classification, so that a user can train
+    it to recognise mails similar to a training set.
+
+    as <a href=http://www.mail-abuse.org>mail-abuse.org</a>, <a
+    href=http://www.ordb.org>ordb.org</a> or others.
+
   - **distributed hash databases**: <A href=http://razor.sf.net/>Vipul's
     Razor</a>, Pyzor and DCC are collaborative spam-tracking databases, which
     work by taking a signature of spam messages.  Since spam typically operates