基于全局本体的多数据源模式匹配方法的研究

Research on Method of Multiple Data Sources Schema Matching Based on Global Ontology

随着信息技术的发展,企业内部各个业务系统之间的联系越来越紧密,数据融合和共享也越来越多样化,各业务系统内的多个异构数据源模式之间往往需要两两相互匹配,但工作量很大、耗时,而且不易维护.针对该问题,提出一种基于全局本体的多数据源模式匹配方法.在该方法中,首先,将待匹配的多个模式转换成统一的数据模型—模式本体;然后,将各个转换后的模式本体分别与全局本体进行模式匹配,分别得到元素间的相似关系;最后,根据相似关系的传递性,将上述得到的相似关系进行聚合,得到多个数据源模式两两之间的匹配结果.通过实验验证,该方法具有高的匹配质量,证明了该方法的有效性.

Along with the development of information technology, the relation between the enterprise internal various business systems is more and more close,data integration and sharing becomes more and more diversified,multiple heterogeneous data sources of business systems often require schema matching each other,but it needs a lot of workload, and becomes time-consuming and not easy to maintain. Aimed at the problem,we put forward a novel schema matching method of multiple data sources based on global ontology. In this meth- od, firstly, transform the multiple schemas into a unified data model-schema ontology;then,match each schema ontology with global ontology to get similarity,respectively;finally,with the transitivity of similarity, aggregate above similarity and get pairwise matching resuits between data sources. Through experiment, this method has high matching quality, and proves the effectiveness of the method.