GLAV集成方法中的模式匹配研究

GLAV是一种新的集成方法,它能为信息集成系统的映射关系提供更具表现力的方式.映射表示源模式与全局模式之间的对应关系,主要研究了GLAV集成方法中产生映射规则的模式匹配方法.首先将SUDAI模型转换为模式树,然后基于树进行匹配.匹配包含语言学和结构匹配两个基本阶段,语言学匹配结合了术语关系和约束进行计算,结构匹配结合了类属性和上下文节点两个方面,并成功解决了模式规模差异带来的负面影响.计算的结果可以判断两个信息类是否是一个映射候选,经过验证后可确定它是否是一个合理的映射.最后定义了一个基本DTD来描述匹配的最终结果.

Schema matching in GLAV integration approach

GLAV is a new integration approach to provide a more expressive way for the mapping of an integration system. Mapping denotes the correspondences between source schema and global schema. This paper mainly studied a schema matching approach in GLAV integration system. The approach changes the SUDAI model to schema tree firstly, and then matches schemata. Matching contains linguistic matching and structure matching. Linguistic matching combines term relation and constraint, while structure matching combines the attributes of a class and context nodes. It has successfully settled negative effect brought by the difference of schema scale. The computing result can judge whether two classes are a mapping candidate or not, and then the users can confirm whether it is a correct mapping by validating. Finally the paper has defined a basic DTD to describe the final result of matching.