基于灰色理论的整体模式匹配  

Gray-theory based holistic schema matching

作　　者：靳光俊[1] 范学峰[1] 郭文宏[1] 金玉[1] 

机构地区：[1]同济大学电子与信息工程学院,上海200092

出　　处：《信息技术》2008年第11期41-45,共5页Information Technology

摘　　要：模式匹配是Deep Web研究中的关键问题。现有的模式匹配算法都是基于两两匹配的原则,同一时刻只能比较两个属性,比较繁琐,当数据源很大时困难很大,效率不高。利用整体模式匹配的思想,提出了一种新的匹配算法。这个方法引入了灰色理论,将每个接口属性抽象成一个向量,每一个向量表示该属性在某一个模式中出现与否。通过比较两个向量各个分量的值,一次性地处理所有模式,从而从整体上分析属性之间的关联性。实验证明这种方法易于计算机实现,可以简单、高效地完成模式匹配。Schema matching is a key problem in Deep Web research. The present schema matching algorithms are all based on pairwise comparison of two interfaces, which are too fussy and, as the amount of data source becomes bigger and bigger, they would become laboured and not efficient. This paper presents a new schema matching algorithm using holistic schema matching idea, which introduces grey theory in the schema matching field, considers every interface property as an abstract vector, each weight represents whether the property appears on certain interface or not. By comparing every weight of two vectors, it can deal with all the schemas once and for all, and analyzes the relevancy of two properties. The experiments prove that this method is easy to realize by computer and can simply and efficiently finish schema matching.

关 键 词：灰色理论 查询接口 模式匹配 关联性 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]