The Bayesian module provides the logic that implements a spam filter based on Bayesian filtering techniques. The filter is trained via the user's own posts (they are considered "safe" content) and the negative food is provided via the posts marked as spam by users via the admin interface.
