Hands-On Artificial Intelligence for Cybersecurity

Advantages of Naive Bayes for spam detection

The aforementioned features are well suited to the task of spam detection. Without the need to resort to large datasets, in fact, the spam detection algorithms based on Naive Bayes can exploit the emails already present in the inbox, constantly updating the probability estimations on the base of new email messages that are progressively added to those already existing.

The constant process of updating the probability estimates is based on the well-known Bayes rule:

The preceding equation describes the relationship between the probability of the occurrence of an event, A, conditioned to the evidence, ...