Paul's Mailfrom (Was: IETF SMTP Working Group Proposal at smtpng.org)

Tue Aug 27 21:54:11 UTC 2002

At 12:58 PM +0100 2002/08/27, michael.dillon at radianz.com wrote:

>  It might be possible to make filters that don't need to be updated that
>  often if
>  you apply AI techniques to recognizing SPAM. For instance, check out this
>  new approach:
>   http://www.paulgraham.com/paulgraham/spam.html

	Bayesian techniques are neither AI nor new.  See 
<http://citeseer.nj.nec.com/sahami98bayesian.html> and 
<http://citeseer.nj.nec.com/androutsopoulos00learning.html>, among 
others.

	Moreover, while there may be client-side programs which implement 
these techniques, and there may be prototype server-side programs 
which implement these techniques, I am not aware of any more 
production-oriented tools which are available to implement these 
kinds of techniques.

-- 
Brad Knowles, <brad.knowles at skynet.be>

"They that can give up essential liberty to obtain a little temporary
safety deserve neither liberty nor safety."
     -Benjamin Franklin, Historical Review of Pennsylvania.

GCS/IT d+(-) s:+(++)>: a C++(+++)$ UMBSHI++++$ P+>++ L+ !E W+++(--) N+ !w---
O- M++ V PS++(+++) PE- Y+(++) PGP>+++ t+(+++) 5++(+++) X++(+++) R+(+++)
tv+(+++) b+(++++) DI+(++++) D+(++) G+(++++) e++>++++ h--- r---(+++)* z(+++)