Generic Schema Matching in IDX Middleware Using Ontologies

It is useful for decision makers to extract web (external) data in order to analyze and promote business activities. But problem with World Wide Web is it has bulk of pages published by variety of end users whose intention is focused on browsing and not for integration. So the problem of mapping unstructured/semi structured web pages data/schema with structured internal warehouse data/schema with no human intervention has become important. This paper proposes a generic composite schema-matching algorithm for mapping external web page schema with internal schema in Intelligent system for data eXchange (IDX), a middleware architecture using broader set of techniques than past approaches. The paper also compares various past approaches for schema matching with IDX schema matcher. At present proposed schema matcher is tested on sample XML Web pages. Test results proved that the whole process is automatic & accurate (if both internal & external data are at same node level. Other wise mapping is done to a higher order level node).