摘要
相似度传播在本体概念相似度计算中有着非常重要的作用。然而,目前常见的相似度传播算法大都采用了固定比例的相似度传播值,没有对相似度转播值进行合理的定量分析。针对此问题,提出了基于概念信息量的相似度传播算法,该算法根据匹配节点的概念信息量大小来判断其子父节点匹配概率大小,通过匹配概率大小调整相似度传播值,从而进行更精确的相似度传播。理论分析与实验结果证明了该算法是有效的。
Similarity propagation is very important for calculating similarity between two concepts. However, the existing algorithms of similarity propagation usually use a fixed proportion of spreading value, these algorithm do not take reasonable quantitative analysis for spreading value. To solve the problem, a novel algorithm of similarity propagation was proposed,which is based on information content of concept. The algorithm adopts the value of information content of matched node to determine matching probability of the matched node's children and parents, and rr^re accurate propagated value will be obtained by adjusting spreading value according to the matching probability of node. Theoretical analysis and the results of experiment show that the algorithm is efficient.
出处
《计算机科学》
CSCD
北大核心
2009年第6期174-177,共4页
Computer Science
基金
国家自然科学基金重点项目(60433020)
湖南省自然科学基金(06JJ50142)
湖南省国土资源厅科技计划项目(200718)资助
关键词
本体
相似度传播
概念信息量
Ontology, Similarity propagation, Information content