基于本体概念的柔性相似度计算方法研究

Study on Method of Flexible Similarity Calculation Based on Ontology Concept

本体可以提供强大的知识表示方法,是信息检索领域中的重要内容。传统的本体概念相似度计算方法大多采用特定于描述语言的通用推理服务来进行匹配,这些方法忽略了概念的语义信息。通过设计一个基于OWL本体的语义检索模型,介绍了如何通过概念的属性以及层次关系来表达概念的语义,计算概念间的柔性相似度。实验结果表明,该方法能充分利用OWL属性特征与层次关系来计算相关概念之间的柔性相似度,可以根据需要动态地调节匹配范围,并给出其在文本分类中的应用。

Ontology can provide a powerful representation of knowledge and solve many semantic problems. It is wonderful to apply on- tology to information retrieval. Traditional concept of ontology similarity calculation methods use general reasoning services of specific description language to match, and these methods ignore semantic information of concept. In this paper by designing an OWL based re- trieval model,introduce how to express semantc of concept and calculate flexible similarity between concepts through concept property and layer relation. The results of experiment prove that the method can fully use OWL property feature and layer relation to compute flexi- ble similarity between concepts, dynamically adjust matching range according to demand, and give applications in the text categorization.