针对现有本体概念相似度计算方法无法处理概念特征属性的模糊性问题,基于本体的语义结构特点提出了通过语义扩展集来对概念特征进行描述的方法,并在此基础上定义概念的模糊特征集合.在特征的量化过程中,充分考虑了本体概念相似度的不对称性、概念的层次深度及区域密度对相似度的影响,最后,通过模糊集合之间的相似度来衡量本体概念的相似度.在通用数据集上的实验表明,该方法的相似度计算性能要优于以往的计算方法.
For the existing ontology concepts similarity measuring methods cannot describe the fuzzy features of concept,based on the semantic structure information of ontology,a method which describes the features of the concept by semantically expanded set is proposed,then a fuzzy set is defined based on the semantically expanded set.As quantifing the features,the algorithm considers influence of asymmetry,depth and local density of ontology concepts on similarity computation.Finally,the semantic similarity between two fuzzy sets is used to measure the concepts similarity of ontology.The result of experiment on a commonly used dataset indicates that our method outperforms traditional similarity measures.