How does Praetor provide better Bayesian
CMS has over 20 years of experience in the world of email and related issues. Drawing from this vast knowledgebase and experience, CMS
designed Praetor to offer better Bayesian filtering.
Upon installation, Praetor has access to a default token database trained from several thousand
message samples. With 85% of the sample messages in this database being
spam, there is no need to perform any initial bulk training.
Praetor's Bayesian filtering protection begins immediately.
From this default starting point, subsequent training can be performed. Praetor's
simplifies this process by automatically capturing samples. You
can then review them to identify which are non-spam messages which the
Bayesian filter classifies as 'Unsure'. For further
recommendations read the Bayesian
Training Tips page.
As your actual message traffic is analyzed and added to the
token database, Praetor's Bayesian filter will work even better.