How does Naïve Bayes connect to Bayesian network factorization constraints and d-separation?

Naïve Bayes is a special Bayesian network where all features are conditionally independent given the class node, enforced by d-separation.

In a Bayesian network, d-separation defines conditional independence relationships. Naïve Bayes is a star-structured graph where the class node is the parent of all feature nodes. This structure guarantees that all features are conditionally independent given the class, simplifying joint probability factorization into P(C)∏P(Xi|C). This makes inference exact and linear in number of features.