DSPAM (as in De-Spam) is an open-source project to create a new
kind of anti-spam mechanism, and is currently effective as both a
server-side agent for UNIX email servers and a developer's library
for mail clients, other anti-spam tools, and similar projects
requiring drop-in spam filtering.

The DSPAM agent masquerades as the email server's local delivery
agent and filters/learns spams using an advanced Bayesian
statistical approach (based on Baye's theorem of combined
probabilities) which provides an administratively
maintenance-free, easy-learning Anti-Spam service custom tailored
to each individual user's behavior. Advanced because on top of
standard Bayesian filtering is also incorporated the use of
Chained Tokens, de-obfuscation, and other enhancements. DSPAM
works great with Sendmail and Exim, and should work well with
any other MTA that supports an external local delivery agent
(postfix, qmail, etc.)