Applying ontologies in the integration of heterogeneous relational databases
Adriana S. Aparício, Oscar Luiz Monteiro de Farias, Neide Santos · 2005
Interoperability is a key point for integrating heterogeneous computing systems. A usual approach proposes the integration of the conceptual databases schemes into a global conceptual scheme, to resolve syntactic and structural heterogeneity. Moreover, semantic problems can remain. Ontologies have been largely designated to overcome semantic heterogeneity. We propose the specification of a formal ontology about the specific knowledge domain to be shared among several database systems build a posteriori of the ontology specification. We reach this integration through a global scheme, developed as a software layer among the databases under consideration. To test this approach we elaborated a case study, based upon hypothetical queries submitted to heterogeneous databases, with data on soil domain, to identify the soil most appropriate to a certain culture. The results are promising, but crucial, in our approach, is the acceptance for a given community of a common vocabulary and its relationships and that are captured by the ontology and transformed to the target conceptual models.