Learning Disjointness Axioms With Association Rule Mining and Its Application to Inconsistency Detection of Linked Data

Disjointness between two concepts is useful to discover new information and detect inconsistencies in knowledge bases. Association rule mining, as a way to discover implicit knowledge in massive data, has been applied to learn disjointness axioms. In this paper, we first analyse the existing method to learn disjointness axioms using association rule mining. Based on the analysis, we propose an improvement of association rule mining for learning disjointness axioms. We then apply the learned disjointness axioms to inconsistency detection in DBpedia and Zhishi.me.