Independence Semantics for BKBs

Bayesian KnowledgeBases (BKB)are a rule-based probabilistic modelthat extend BayesNetworks(BN), by allowing context-sensitive independenceand cycles in the directed graph. BKBshaveprobabilistic semantics, but lack independencesemantics, i.e., a graphbased scheme determining what independence statements are sanctioned by the model. Such a semantics is provided through generalized dseparation, by constructing an equivalent BN. While useful for showingcorrectness, the construction is not practical for decision algorithms due to exponential size. Someresults for special cases, whereindependence can be determined from polynomial-time tests on the BKBgraph, are presented.

• Type: Conference
• Year: 2000
• Where: FLAIRS
• Authors: Solomon Eyal Shimony, Eugene Santos Jr., Tzachi Rosen
• Keywords: Artificial Intelligence, Bayesian KnowledgeBases, Context-sensitive Independenceand Cycles, FLAIRS 2000, Rule-based Probabilistic Modelthat