计算机应用

• 人工智能 • 上一篇    下一篇

基于启发式规则的本体概念语义相似度匹配

杨哲   

  1. 苏州大学计算机科学与技术学院
  • 收稿日期:2007-06-25 修回日期:2007-08-16 发布日期:2007-12-01 出版日期:2007-12-01
  • 通讯作者: 杨哲

Semantic similarity match of ontology concept based on heuristic rules

Yang Zhe   

  • Received:2007-06-25 Revised:2007-08-16 Online:2007-12-01 Published:2007-12-01
  • Contact: Yang Zhe

摘要: 在本体概念模型的层次结构中,概念的语义相似度与概念间的语义距离成反比例关系,同时越靠近底层的概念所描述的信息越具体,因此若概念间最近共同祖先的深度越大,则概念的语义相似度越大。考虑到这两个因素,构造了计算本体概念语义相似度的启发式规则及相应的公式,通过对本体实例的分析,验证了公式的有效性。公式中两个经验参数的选择与本体概念模型的深度有关,根据目前本体概念模型的深度大小,给出了经验参数合理的取值。

关键词: 本体, 概念模型, 语义相似度, 启发式规则

Abstract: In the hierarchy model of ontology conceptual model, the semantic similarity between two ontology concepts is in inverse proportion to the semantic distance in between. At the same time, the closer the concepts are to the bottom in the hierarchy model, the more information they describe. Therefore, if the deeper the nearest shared ancestor is, the more similar the two concepts are. Considering these two factors, two heuristic rules and the calculating formula were constructed. And the formula was proved reasonable by analyzing an ontology instance. In the formula, there were two experimental parameters, that were decided by the depth of the hierarchy model of ontology. According to the depth at present, the reasonable values of two experimental parameters were given.

Key words: ontology, concept model, semantic similarity, heuristic rule

中图分类号: