Bayesian networks is a probabilistic graphical model representing a set of variables and their conditional dependencies via a directed acyclic graph (DAG). This academic concept and methodological approach investigates the structure of probabilistic relationships among variables, where nodes in the DAG represent variables and directed edges indicate direct dependencies. Key characteristics include the qualitative representation of dependencies through the graph structure and the quantitative specification via conditional probability distributions associated with each node. Their significance lies in providing a framework for probabilistic inference, learning from data, and supporting decision-making under uncertainty across various domains.