基于概念格的跨本体映射中概念相似度计算方法

An Approach to Concept Similarity Computation in Cross-Ontology Mapping Based on Concept Lattice

随着各种本体构建方法和诸多实验本体的相继涌现,以跨本体通信、跨本体协同为目的的本体对应相关研究在近年来受到国际学术界的普遍关注.为了最大限度复用现有本体,解决本体对应中跨本体映射的核心问题,在对当前本体映射中概念相似度的计算方法进行梳理和总结的基础上,提出了以＂基于概念格的对象-属性相似度（object-attribute similarity based on concept lattice,OASBCL）＂法计算跨本体映射中概念的相似度.通过对该方法在跨本体映射中的应用举例,阐明了方法的有效性.并在此基础上从概念格与本体互补、相似度要素指标、映射性质三个方面对该方法进行了讨论.以尝试探索一种能够支持异构本体间跨本体映射的形式化的概念相似度计算方法.

With multiple methods of ontology construction and many experimental ontologies have emerged, the study of ontology alignment was paid attention in international academia for ontology correspondence and ontology interoperability. In order to reuse the existing ontologies and solve the core issues of cross-ontology mapping in ontology alignment, based on the summary of methods for concept similarity computation in ontology mapping, this article proposes the ＂Object-Attribute Similarity Based on Concept Lattice （OASBCL）＂ to compute the concept similarity in cross-ontology mapping. By the example of the method in a cross-ontology mapping application, the effectiveness of the method is illustrated. And the method is discussed on three areas, complementariness of ontology and concept lattice, factors of similarity, character of mapping. The goal of this article is to explore a way to compute the concept similarity formally in order to support crossontology mappings between heterogeneous ontologies.