Instance based Matching using Regular Expression

Abstract

Instance based matching is the process of comparing data from different heterogeneous data sources in determining the correspondence of schema elements. It is a useful alternative choice when schema information (element name, description, constraint) is unavailable or unable to determine the match between schema elements. Instance based matching is a non trivial problem and is applied in many application areas such as data integration, data cleaning, query mediations, and warehousing. Many instance based solutions to the schema matching problem have been proposed and most of them utilized similarity metrics. In this paper, we present a fully automatic approach that contributes to the solution of instance based matching in identifying the correspondences of attributes which is one of the elements in the schema by utilizing regular expression. Several experiments using real-world data set have been conducted to evaluate the performance of our proposed approach. The results showed that our proposed approach achieved better accuracy compared to previous approaches using similarity metrics.