基于本体的概念相似度计算的改进  

Improved Ontology-based Concept Similarity Computation

作　　者：崔春华[1] 李华[1] 

机构地区：[1]重庆大学计算机学院,重庆400044

出　　处：《世界科技研究与发展》2013年第2期211-213,共3页World Sci-Tech R&D

基　　金："基于本体;语义和语用的智能化教育资源应用平台"横向课题;国家社科基金"十一五"规划教育学2007年度国家重点课题(ACA070);重庆大学承担子课题八(ACA0704-08)资助

摘　　要：通过对概念相似度计算方法分析研究,提出一种改进的基于本体的概念相似度计算模型。该模型综合考虑了语义距离、关系类型、节点密度、语义深度、语义重合度及概念属性等因素对概念相似度计算的影响,且可以根据不同领域本体进行参数调节。实验结果表明,该模型显著提高了概念相似度计算的准确性,所得相似度值与人类主观判断曼为接近。Through analyzing and studying the methods of concept similarity computation, an improved ontology-based concept similarity computation model is proposed. The factors of influencing semantic similarity computation are synthetically considered, such as semantic distance, relationship type, node density, semantic depth, concept attribute, semantic coincidence degree. Parameters are adjusted according to the different domain ontology. The experimental result shows that the accuracy of concept similarity computation is obviously enhanced, with the similarity value obtained much closer to human subjective judgments.

关 键 词：本体 概念相似度 语义距离 

分 类 号：TP391.1[自动化与计算机技术—计算机应用技术]