Semantic similarity analysis of XML schema using grid computing

Abstract

A growing number of e-businesses have been using XML schemas in recent years. Schema mapping now plays a crucial role in integrating heterogeneous e-business applications. Since large-scale XML schema mapping using complex and hybrid similarity measures requires significant amount of processing time, a sophisticated similarity analysis algorithm is needed to handle its complexity and performance. In this paper, we focus on designing a service-oriented architecture (SoA) for schema mapping, based on a grid computing technology in order to enhance the effectiveness of the mapping algorithm. After comparing three different grid computing technologies (MPJ, Hadoop, and Globus), we explain why MPJ is the most suitable. We propose SoA XML schema mapping based on MPJ, and demonstrate its performance.