基于核矩阵优化方法的本体算法  

Ontology Algorithm Based on Kernel Matrix Optimization Method

作　　者：彭波[1] 徐天伟[1] 李臻[1] 高炜[1] 

机构地区：[1]云南师范大学信息学院,昆明650500

出　　处：《科学技术与工程》2013年第26期7692-7696,共5页Science Technology and Engineering

基　　金：国家自然科学基金(60903131);教育部科学技术研究重点项目(210210)资助

摘　　要：本体相似度计算和本体映射被广泛应用于查询扩展和图像检索中,已成为信息科学研究的热点内容,其核心为计算本体图中顶点间的相似度。用核矩阵表示本体图中每一对顶点的相似度,根据相邻顶点相似度大、不相邻顶点相似度小的特征,结合转换函数的光滑性得到核矩阵优化模型,求解模型得到最优核矩阵。将此方法分别应用于生物GO本体和数学学科本体,通过实验表明新本体相似度计算和本体映射算法有较高的效率。Ontology similarity computing and ontology mapping are widely used in query expansion and image retrieval. The core trick for ontology applications is calculating the similarity between vertices in ontology graph, and it has become a hot topic in information science research. The similarity for each pair of vertices can be express as kernel matrix. The optimization model is given according to the fact that adjacent vertices with higher similarity and non-adjacent vertices with lower similarity, and combined with the smoothness of the transfer function. The optimal kernel matrix is obtained by solving this model. This method were employed in biological GO ontology and mathematics ontology, and experiments results show that the new ontology similarity calculation and ontology mapping algorithms have higher efficiency.

关 键 词：本体 相似度 本体映射 光滑性 核矩阵 

分 类 号：TP393.092[自动化与计算机技术—计算机应用技术]