Authors:
Amar Zerdazi
and
Myriam Lamolle
Affiliation:
LINC – University of Paris VIII, IUT of Montreuil, France
Keyword(s):
Heterogeneity, schema matching, semantic integration, XML schema.
Related
Ontology
Subjects/Areas/Topics:
Internet Technology
;
System Integration
;
Web Information Systems and Technologies
;
XML and Data Management
Abstract:
Schema matching is a critical step in integration of heterogeneous data sources. Recent integration work has mainly focused on developing matching techniques to find equivalent elements among the different XML sources. In this paper we propose a new approach to structural similarity measure based on the notion of context, between entities of the Enhanced XML Schemas, called EXS. In our approach, the set of the EXS schemas, are considered like a federation of XML schemas descended of different heterogeneous sources schemas (relational, object, XML, etc.) and enriched by the semantic metaknowledge. We present here the major problems bound to this crucial task, notably with regard to the semantic of schemas. So, we propose a structural matching algorithm. The algorithm takes two schema graphs as input, and produces as output a mapping between corresponding nodes of the schema graphs. After our algorithm runs, we expect a human to check and adjust the results.