Sreyanth writes:

I have implemented a binary Bayesian classifier which classifies an email either spam or not spam.

Is it better than (or at least different from) SpamBayes (available on PyPI), which some third parties already use in Mailman installations? Let's not reinvent wheels, here!