Abstract:
The account quantum of ontology mapping based on similarity computation is vast, because the amount of concepts and attribute waiting for computing is large. This paper adopts a filtering tactic which uses candidate mapping tactic and information gain tactic to reduce the amount of concepts and attribute waiting for computing. This filtering tactic makes full use of the characteristic of ontology and the idea of data digging, and gets rid of those worthless concepts and attribute to reduce the account quantum of similarity computation. Experimental results indicate that the effect of those concepts and attribute is little.
Key words:
ontology mapping,
candidate mapping,
information gain
摘要: 在基于相似度计算的本体映射中,相似度计算量大的主要原因是待映射概念和待计算属性过多。该文采用过滤策略,利用候选映射策略和信息增益策略减少待映射概念和待计算属性数量。该过滤策略充分利用本体特点和数据挖掘思想,有效滤除没有计算意义的概念和属性,减少了相似度计算量。实验结果证明,滤除的概念和属性对映射效果的影响很小。
关键词:
本体映射,
候选映射,
信息增益
CLC Number:
GU Zhi-feng; LIU Yong; GUO Gen-cheng. Optimizing Method for Ontology Mapping Based on Similarity Computation[J]. Computer Engineering, 2008, 34(19): 56-57,6.
谷志锋;刘 勇;郭跟成. 基于相似度计算的本体映射优化方法[J]. 计算机工程, 2008, 34(19): 56-57,6.