An approach for developing a schemaless XML dataspace profiling system

Abstract

Dataspace systems have been proposed recently as an alternative to the traditional data integration approach. They offer reduced setup time and costs by allowing cooperation among the data sources based on the knowledge gained gradually through user interaction. A dataspace is a collection of heterogeneous and beforehand unfamiliar but interrelated data sources. In this article, we consider dataspaces composed of XML-based data sources. The XML query and analysis systems designed to satisfy the user’s sophisticated information needs presuppose that they are familiar with the contents, structures and semantics of the underlying data sources. In order to provide this information, we introduce and specify a schemaless XML dataspace profiling system that assists the user in selecting data sources relevant to him/her and in validating their consistency by detecting the potential data conflicts among them. We also demonstrate how our approach affords the possibility of utilizing an advanced XML query system.