基于 DL 关联εL ＋＋规则挖掘的归纳知识发现.

In order to discover knowledge from incomplete and dynamic data, this paper proposed a technique for consistent knowledge discovery based on DL association ε L+ + rules and inductive reasoning. Firstly, by analyzing description logic sL + + rules and the dynamics of the knowledge of evolving ontology, it obtained inductive reasoning learning in evolving ontology. Basing on the support and weight of atomic set and the confidence of association sL + + rules, it achieved ε L+ + rules mining. Secondly, by getting the representative association DL sL + + rules with minimum support and minimum weight, it realized the precise identification of fundamental rules so as to complete inductive knowledge discovery. The experimental results for historical data from a certain city show that, compared with the existing mainstream methods, the proposed technique not only has better scalability, but also has higher accuracy in terms of evolving ontology and dynamic semantic data knowledge discovery. [ABSTRACT FROM AUTHOR]