摘要
概念的语义相似度研究,是知识表示以及信息检索领域中的一个重要内容。通过分析两种传统的语义相似度计算方法,对它们存在的问题进行改进,提出了一种综合的基于本体的概念语义相似度计算方法。该方法结合本体网络特征和语义距离计算中的多种语义影响因素,充分利用本体中概念的语义信息计算概念间的语义相似度。实验结果比较合理,验证了该方法的有效性。
Research on semantic similarity of concept is very important in knowledge representation and information retrieval. By analyzing two traditional semantic similarity computation methods and modifying the problems in them, this paper puts forward a cornpositive method based on ontology. The method integrates the graph structure of ontology with several kinds semantic influence factors in computation of semantic distance, and utilizes the concept semantic information in ontology fully to compute the semantic similarity between concepts. The experiment result shows that the computation method is reasonable, its effectiveness is verified as well.
出处
《计算机应用与软件》
CSCD
2009年第7期143-145,共3页
Computer Applications and Software
关键词
本体本体
语义相似度
语义距离
信息量
共同祖先
Ontology
Semantic similarity
Semantic distance
Information content
Common ancestor
