Weka is a collection of machine learning algorithms for data mining tasks. The algorithms can either be applied directly to a dataset or called from your own Java code. Weka contains tools for data pre-processing, classification, regression, clustering, association rules, and visualization. It is also well-suited for developing new machine learning schemes.

Look at the Bayesian Network Classifiers, FAQ at:


There is also a bigger comprehensive documentation at:


Source: Weka FAQ