We define ontology as the formal specification of a vocabulary of concepts, including axioms relating those concepts. Database Integration is the problem of answering queries via a unified view over a set of possibly heterogeneous sources. The difficulty lies in the preservation of semantics between and across sources. Inferential integration uses ontologies to reduce this difficult problem to that of sound inference.
OntoGrate is a prototype system implementing this idea. We have implemented two heterogeneous databases (Nwind* and Stores7**) from the similiar domain of e-commerce (customers, orders, items, etc.) The schema diagrams are available for Nwind Schema and Stores7 Schema. Also, the ontology representation of those schemas are available for Nwind Ontology and Stores7 Ontology in our internal language called Web-PDDL. The mapping between them (providing a unified view) is called a merged ontology and is also available for Stores7-Nwind.
Please try OntoGrate below.
You also can try another online service of ontoGrate, which queries Relational Database with Semantic Web Ontologies.
For more information, please read our publications referenced below.Back to top «
OntoGrate is flexible. Suppose you are a database user from the Nwind system, you can issue a query using your knowledge of Nwind's schema to find answers from Stores7 (and vice-versa). Alternatively, you can choose to use the more general (merged) ontology to specify your queries and get answers from both databases.Back to top «