Data Integration: Viewpoint Based Approach

Nowadays, the mediation of information from multiple autonomous and heterogeneous data sources becomes a crucial task for future Web information systems. With the succession of ontologies, an important progress has been realized in the automation of the integration process seen their explicit representation of data semantics. Somewhere else, the paradigm of viewpoints can be an undeniable contribution in the integration of large number of information sources by several viewpoints. In this paper we propose an hybrid approach of integration based on the notion of viewpoints which takes into account new hypotheses: (1) data sources are grouped according to the viewpoint that they represent, (2) mediators are represented in two different levels of the architecture, and (3) an hybrid approach that consists in combining the two data integration approaches Global As View (GAV) and Local As