• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

J4 ›› 2010, Vol. 32 ›› Issue (10): 93-96.doi: 10.3969/j.issn.1007130X.2010.

• 论文 • 上一篇    下一篇

高效的异构本体的映射算法研究

王松1,2,马勇1,王刚1,刘晓光1   

  1. (1.南开百度联合技术实验室,南开大学信息学院,天津 300071;2.装备保障系军事交通学院,天津 300161)
  • 收稿日期:2010-03-17 修回日期:2010-06-20 出版日期:2010-09-29 发布日期:2010-09-29
  • 作者简介:王松(1983),男,天津人,本科,研究方向为基于本体的P2P语义网络检索;马勇,本科,研究方向为P2P语义网络检索系统;王刚,博士,教授,研究方向为分布式与并行计算、网络存储;刘晓光,博士后,副教授,研究方向为分布式与并行计算、网络存储。
  • 基金资助:

    国家863计划资助项目(2008AA01Z401);国家自然基金资助项目(60903028);教育部博士点基金资助项目(20070055054);天津市科技发展计划资助项目(08JCYBJC13000)

An Efficient Heterogeneous Ontology Mapping Algorithm

WANG Song1,2,MA Yong1,WANG Gang1,LIU Xiaoguang1   

  1. (1.NankaiBaidu Joint Lab,School of Information Science,Nankai University,Tianjin 300071;
    2.Equipment Support Department,Military Transportation University,Tianjin 300161,China)
  • Received:2010-03-17 Revised:2010-06-20 Online:2010-09-29 Published:2010-09-29

摘要:

基于本体的概念间相似度计算已经在信息检索等诸多领域成为当今信息技术研究的热点问题之一。本文的工作是针对描述同一领域的多个本体间存在的异构问题,设计一种快速高效的映射算法来实现异构本体的融合。本文提出了一种基于异构本体的相似度计算方法,通过字面概念相似度和语义结构(包括节点深度、节点密度、边权重、信息量等)相似度等方面的综合计算,可以准确地得到异构本体间的概念映射关系;同时,通过对映射方法的优化,算法的匹配速度也有很大程度的提高。实验结果表明,该算法可以有效地排除本体异构的影响,得到较好的概念相似性计算效果

关键词: 异构本体, 概念相似度, 映射

Abstract:

The research of ontologybased similarity calculation between concepts has already been a hot issue of information technology in the fields of information retrieval, and so on. In this paper, the contents of the study is to find a fast and efficient mapping algorithm for heterogeneous ontologies in the same field.This paper puts forward a method of similarity calculation based on heterogeneous ontologies,and the factors of similarity of literal meaning and semantic structure(including the depth of the node, node density, edge weight,information content, etc.) can get concept mapping between heterogeneous ontologies more accurately.Simultaneously,taking into account the optimization of mapping method,the speed of matching has also been improved to a large extent. The problem of how to improve the speed of matching more effectually has been mentioned in this paper.The experimental results show this method can effectively get better effectiveness with concept similarity computing, excluding the effects of heterogeneous ontologies.

Key words: heterogeneous ontology;concept similarity;mapping