A disciplined approach for the integration of heterogeneous XML datasources

In this paper, we focus on the problem of integrating heterogeneous XML datasources. We follow a semantic approach to information sharing and integration, and we present a disciplined approach based on reconciliation rules and mapping rules to set up a mediation scheme describing the information about DTDs and their contents at an ontological level, without altering the original format of the XML data. The mediation scheme is an XML-based description of the relevant concepts, relationships and properties featuring heterogeneous DTDs. It is derived from a semantic matching of DTD contents, and is organized as a network of concepts and links, with associated mapping rules to the underlying XML datasources.