-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
It's SVMs -- another type of probabilistic classifier, and probably requires training. A well-trained filter gets high accuracy, but training it is hard work :( - --j. Chris Santerre writes: > >-----Original Message----- > >From: Kang , Joseph S. [mailto:[EMAIL PROTECTED] > >Sent: Tuesday, August 24, 2004 2:23 PM > >To: Spamassassin-Users ([email protected]) > >Subject: [OT] Spam FIREWALL software > > > > > >Heard about this via /. and followed my nose to the Univ. of > >Queensland site > >for the official press release. > > > >http://www.uq.edu.au/news/index.phtml?article=5833 > > > >Revolutionary spam firewall > >Monday , 23 August 2004 > > > > > *snip* > > Sounds like bayes. > I heard loud and clear "seeking investment" > > And most impotant of all, it does nothing to stop spam beyond its own > boarders. > > "Existing anti-spam software filters out spam whereas ours puts up a > firewall, stopping all email traffic and only allowing real mail through," > said Mr Sullivan. > > Oh my!! Well, we could all do that in 5 minutes. Gee I wonder if that is a > good idea? So kind of like an RBL only before the end of the data command? > Hmm....gee....never heard that before :) > > Or does it process the whole thing, then what? /dev/null? > > I'm starting to sound more jaded everyday :) But I see absolutely nothing > new here at all. Maybe one of the SA devs knows more? And if that is his > real world FP rate, then I'm the lost Anastasia of Russia! > > --Chris -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Exmh CVS iD8DBQFBK5MoQTcbUG5Y7woRAuZVAKC8+axzSDiUEOYZj8DwxAO6rvIazQCgjrAE V4kUtp17nNUJlDM6dSK7OMA= =lBOz -----END PGP SIGNATURE-----
