[ Date Index ][
Thread Index ]
[ <= Previous by date /
thread ]
[ Next by date /
thread => ]
On Tuesday 13 April 2004 23:54, Dave Trudgian wrote:
The problem I have is huge spam volumes, some of which is very simiar to genuine email, thus the majority of what the Bayesian and similar filters let through is still spam.Yeah, this is a big problem in spam classification. You have to be able to learn very subtle differences, without losing the ability to handle massive differences between blatant spam and ham mail. It's all about identifying the right features to classify on.
What is amazing is how quickly people can do it though, isn't it. We only need to look further than the subject in maybe 10% of messages that make it through the Bayesian filter and other protections, and even when we do it takes only moments. Wetware rules! So to solve the spam problem, first, solve the AI Problem. -- Adrian Midgley (Linux desktop) GP, Exeter http://www.defoam.net/ -- The Mailing List for the Devon & Cornwall LUG Mail majordomo@xxxxxxxxxxxx with "unsubscribe list" in the message body to unsubscribe.