Title :
A domain knowledge uncertainty reasoning method combining Bayesian network with conditional event algebra
Author :
Dong, Rensong ; Wang, Hua ; Yu, Zhengtao
Author_Institution :
Sch. of Metall. & Energy Eng., Kunming Univ. of Sci. & Technol., Kunming, China
Abstract :
Aiming at difficult problems of Inconsistent of probability and logic expressions and high-end probability logical reasoning for solving domain knowledge. This paper proposes a domain knowledge uncertainty reasoning method with a combination of Bayesian network and conditional event algebra. This method used Bayesian formula combining with the conditional independence assumption to do causal reasoning existing in a variety of graphs according to the high efficiency of expressing uncertainty reasoning by Bayesian network, and transformed a higher-order conditional event to normal event solution via Condition Event Algebra, so the uncertainty reasoning solution of domain knowledge can be realized. Through analyzing an uncertain reasoning case solving based on the tourism domain ontology, the results show that the proposed method is effective, and can well solve approximate reasoning problems of domain knowledge uncertainty reasoning.
Keywords :
belief networks; graph theory; inference mechanisms; ontologies (artificial intelligence); probabilistic logic; process algebra; uncertainty handling; Bayesian formula; Bayesian network; approximate reasoning problem; causal reasoning; conditional event algebra; conditional independence assumption; domain knowledge uncertainty reasoning method; graphs; high-end probability logical reasoning; logic expressions; tourism domain ontology; Algebra; Bayesian methods; Cognition; Educational institutions; Knowledge engineering; Markov processes; Uncertainty; Bayesian network; Condition Event Algebra; Domain knowledge; Uncertainty reasoning;
Conference_Titel :
Control Conference (CCC), 2012 31st Chinese
Conference_Location :
Hefei
Print_ISBN :
978-1-4673-2581-3