中文摘要：

概念相似度研究,是知识表示以及信息检索领域中的一个重要内容.通过对传统相似度计算方法进行分析,提出了一种改进的概念相似度计算模型.该计算模型在计算相似度时不仅改进了语义距离、层次差、语义重合度的计算方法,还考虑了节点密度和有向边类型对相似度计算的影响.实验结果表明,该方法充分利用了本体层次树的结构特点来计算概念之间的相似度,全面地量化了本体概念节点间的语义相似度,提高了概念间相似度计算的准确性.

英文摘要：

The research of concept similarity is an important content in the area of knowledge representation and information retrieval.A new modify model of computing the similarity of concepts can be proposed by analyzing the traditional computing methods of concepts similarity.In the process of computing semantic similarity,the new model considers not only the improved computing methods of the semantic distance,level differences and semantic ratio,but the node density and the type of edges for the effects of similarity computing.The experiment results show that this method makes full use of the level structure of ontology tree to compute the similarity of concepts,fully measures the semantic similarity between two nodes in ontology,and improves the accuracy of computing the semantic similarity of concepts.