摘要： 针对当前概念相似度计算的片面性和不完善性等不足，提出一种改进的基于语义距离的概念间语义相似度计算方法。从有向边包含的信息量、有向边的类型以及概念密度3个方面对语义距离进行扩展，将语义距离转换成语义相似度，通过引入不对称因子，使最终概念语义相似度计算更加精确。将该方法与基于信息量方法、基于距离方法及人的主观判断结果进行比较，验证了该方法的可行性和有效性。

关键词: 本体, 语义相似度, 语义距离, 语义密度, 权重, 不对称因子

Abstract: Aiming at the problem of the one-sided and incomplete nature for the current concept similarity computation, this paper presents an improved concept semantic similarity computation method which is based on semantic distance. It spreads semantic distance from three sides of the information contained in directed edge, the directed edge type and concept density, turns semantic distance to semantic similarity. At the same time, this paper introduces the dissymmetry factor, and makes the last concept semantic similarity computation more exact. Comparing this method with information-based method, distance-based method and the human subjective judgment, result proves that this method is feasible and valid.

Key words: ontology, semantic similarity, semantic distance, semantic density, weight, dissymmetry factor

中图分类号: 

• TP311