Abstract: Aiming at the heterogeneity problem between ontology, an ontology matching method based on Top-k mapping is put forward. This is an extension of existing methods. It takes similarity computation as a basis of the algorithm by calculating the similarity of two concepts at element and structure level. Moreover, it can generate k mapping simultaneously not one best mapping. Experimental shows that the method has advantages in the recall and precision, and excels GLUE in precision.

Key words: ontology, ontology matching, similarity, Top-k mapping

摘要： 针对本体之间的异构问题，提出一种基于Top-k映射的本体匹配方法。该方法是对现有匹配方法的一种扩展，它以相似度计算为基础，从元素级和结构级计算2个概念之间的相似度，并在匹配过程中同时产生k个映射而不是一个最佳映射。实验结果表明，该算法在查全率和查准率方面都有很好的表现，并且其查准率要优于GLUE方法。

关键词: 本体, 本体匹配, 相似度, Top-k映射

CLC Number: 

• TP311