On Thursday February 19 2004 11:51, Jennifer Fountain wrote:
> Does bayes filtering resolve 100% of your spam problems? My pres of IS
> thinks that bayes filtering is the end all that ends all of spam. He
> thinks if we only use bayes we will stop spam entirely.
>
> Can anyone make a statement on this? I may have to buy a commerical
> product because I cannot get spamassassian to do it with my
> configuration - we currently use qmail-scanner.
>
> Thanks for any information.
As you may well know, false-positive is a fundamental fact of spam filtering.
Hence no filtering will give you 100% accuracy. Don't use spam filtering
(commercial or open source) if you are not prepared to deal with losing mail.
The way I justify it is that the probability of SA mis-tagging email is much
lower than that of me sitting there manually scanning through thousands of
messages. Not to mention that saved time and effort.
Although I did manage to get around 99.99% accuracy, with 0.01%
false-positive ... All by using SpamAssassin, with bayes and network tests.
Pedro
--
A rock pile ceases to be a rock pile the moment a single man
contemplates it, bearing within him the image of a cathedral.
-- Antoine de Saint-Exupery
