Basic Principles of Learning Bayesian Logic Programs

• Published in: Probabilistic Inductive Logic Programming Vol. 4911; pp. 189 - 221
• Main Authors: Kersting, Kristian, De Raedt, Luc
• Format: Book Chapter
• Language: English
• Published: Germany Springer Berlin / Heidelberg 2008; Springer Berlin Heidelberg
• Series: Lecture Notes in Computer Science
• Subjects: Artificial intelligence; Bayesian Network; Computer programming / software development; Data Case; Dynamic Bayesian Network; Inductive Logic Programming; Logic Program; Mathematical theory of computation
• ISBN: 9783540786511; 3540786511
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-540-78652-8_7

Bayesian logic programs tightly integrate definite logic programs with Bayesian networks in order to incorporate the notions of objects and relations into Bayesian networks. They establish a one-to-one mapping between ground atoms and random variables, and between the immediate consequence operator and the directly influenced by relation. In doing so, they nicely separate the qualitative (i.e. logical) component from the quantitative (i.e. the probabilistic) one providing a natural framework to describe general, probabilistic dependencies among sets of random variables. In this chapter, we present results on combining Inductive Logic Programming with Bayesian networks to learn both the qualitative and the quantitative components of Bayesian logic programs from data. More precisely, we show how the qualitative components can be learned by combining the inductive logic programming setting learning from interpretations with score-based techniques for learning Bayesian networks. The estimation of the quantitative components is reduced to the corresponding problem of (dynamic) Bayesian networks.