An ontology-based framework for XML semantic integration

XML is becoming the standard for data interchange on the Web. However, XML and its schema languages do not express semantics but rather structure, such as nesting information. Therefore, semantically equivalent documents often present different document structures. We provide an ontology-based framework that aims to make two XML documents intemperate at the semantic level while retaining their nesting structure. In our global-as-view approach, we generate an RDF ontology for each of the participating XML documents, which preserves the nesting structure of the document. An RDF global ontology is the result of merging the individual ontologies. The global ontology unifies the query access and establishes semantic connections among the underlying individual databases. We consider two types of queries: those that are posed on the global ontology and those that are posed on any of the XML documents, in a P2P fashion. The former type is processed using query translation from an RDF query to an XML query. The latter type entails bidirectional query processing: the translation from an XML query to an RDF query followed by the translation from an RDF query to an XML query. To ensure the correctness of the answer to the query in the latter case, we introduce the concept of reversibility of the query translation.