Infinite Monkeys and other BLs

[Peter Bonivart]

Chris W. Parker wrote:
> i've never had a false negative* so i imagine that if i put bayesian at
> the gateway my spam count would go way down at the client level if not
> disappear.

Don't count on it, it's a lot easier to train the Bayesian filter on a
per-user basis than for a whole domain. It will be less accurate but
still most important. It's my best spam trap but I still see it fooled
all the time. For example, spam in other languages than English (Swedish
in my case) gets by easy, luckily most spam is in English.

> * i always get confused on this but.. false negative is when spam is not
> detected as being spam right? and a false positive is when non-spam *is*
> detected as spam?

Yes. :-)

--
/Peter Bonivart

--Unix lovers do it in the Sun

Sun Fire V210, Solaris 9, Sendmail 8.12.10, MailScanner 4.25-14,
SpamAssassin 2.63 + DCC 1.2.30, ClamAV 0.67 + GMP 4.1.2, MailStats 0.25