AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF ENGINEERING
Pagination or Media Count:
Abstract Managing uncertainty in complex domains requires a flexible and semantically sound knowledge representation. This is especially important during the initial knowledge engineering and subsequent maintenance of the knowledge base. We present a new model of knowledge representation called Bayesian Knowledge Bases. It unifies an if then style rules with probability theory. We can prove that such a merger remains fully probabilistic and yet maintains full flexibility and intuitiveness.