一种多策略通用模式匹配方法  

A Multistrategy Generic Schema Matching Approach

作　　者：程伟[1] 周龙骧[2] 林河水[1] 孙玉芳[1] 

机构地区：[1]中国科学院软件研究所,北京100080 [2]中国科学院数学与系统科学研究院,北京100080

出　　处：《计算机科学》2004年第11期121-123,共3页Computer Science

基　　金：国家自然科学基金(19831020)

摘　　要：模式匹配问题即寻找不同模式元素之间的语义对应关系,在数据仓库、异构数据源集成及语义Web等领域都是非常重要的研究基础。目前模式匹配仍大多主要由人工来完成,因此有很大局限性。提出了一种多策略通用模式匹配架构,可以方便地兼容其他匹配策略。采用了一种基于词语间语义距离的方法来计算其语义相似度;提出了一种基于相似度传播的结构匹配算法,有效地考虑了相邻相似节点间的相互影响。实验结果表明这种匹配方法在处理模式匹配任务时能达到较高的精度。Schema matching,the problem of finding semantic correspondences between elements of two schemas,plays a key role in many applications,such as data warehouse,heterogeneous data sources integration and semantic Web. Currently,schema matching is largely performed manually by domain experts,thus a time-consuming and labor-intensive process. In this paper,we describe a multistrategy schema matching framework,which can combine multiple matching strategies flexibly and its architecture is extensible to new marchers. We adopt an approach based on semantic distances between words to compute their semantic similarity. We propose a structural matching algorithm based on semantic similarity propagation,which consider the effect between neighboring nodes. After describe our approach,we present experiment results on several real-world domains ,and show that the approach discovers semantic mappings with a high degree of accuracy.

关 键 词：模式匹配 通用 结构匹配 异构数据源集成 语义相似度 语义WEB 匹配策略 配方法 对应关系 实验结果 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TP311[自动化与计算机技术—计算机科学与技术]