On Wed, Dec 18, 2002 at 01:40:46PM +0000, MJ Ray wrote:
- I can't read Chinese or Korean, so most of that is filtered out. (Sorry.)
1.5 List emails get filtered off here at present. ;-)
- A Distributed Checksum Clearinghouse filter (dccproc) filters off most
bulk mails (nb, that is not quite the same as spam, but close enough for most).
- A slightly under-trained (1500 emails seen in total) Bayesian word
probability filter gets the rest.
I get about 2 spams in my inbox per day, on average, which is a false negative rate of about 0.2%. I haven't had any non-spam non-bulk email in the spamtrap yet.
What methods programs etc. do you use for all of these? homebrewed or does it come from somewhere else? It would be interesting to know as I have been running spamassassin for nearly a week and it gets too many bulk emails that I actually want to receive (and remembering to add everything to whitelist every time I subscribe is a pain) and is still letting through quite a few spams.
I want to look at alternatives anyhow before I go through and actually configure spamassassin a bit more...
Adam