基于机器学习的本体概念相似性研究

Research on Ontology Concept Similarity Based on Machine Learning

现有的知识系统使用的是集中式的、一致性的、可扩充的Ontology库,不同本体间的语义匹配是语义网发展面临的最富挑战性的问题之一。本文针对领域中存在不同的Ontology的问题,讨论了一种基于多策略机器学习的Ontology匹配方法,重点分析了本体概念的相似度计算,并提出了一种相似度测量算法。

Current knowledge system is using a centralized, persistent, scalable ontology repository. Mapping between semantic heterogeneous ontologies among semantic Web is becoming an ever more important issue. In this paper, an ontology mapping system is described, which applies multiple classification methods to learn the mapping between ontologies, and then the similarity measuring method between different concepts which belongs to different ontology respectively is analyzed. As a result, the algorithm of similarity measure is brought forward.