Interoperability and Integration of Independent Heterogeneous Distributed Databases over the Internet

Abstract

A key challenge for independent databases that are distributed over the Internet is to provide mechanisms for interoperability to facilitate resource discovery, access, distributed processing and integration. Additional challenges arise when there is semantic heterogeneity of the underlying ontologies associated with the databases. This paper develops a mechanism that facilitates interoperability of independent distributed database applications that is useful in situations where databases have evolved separately but where post hoc semantic mappings between schema are possible. Such functionality can be obtained by storing metadata in a common format, such as RDF, thus allowing the databases to be discovered and accessed by browsing registries that contain details of the database objects. Once interoperability has been achieved, we describe a flexible user-centric method for integration and knowledge discovery from semantically heterogeneous data, based on the specification of ontology mappings from distributed local data sources to a derived ontology of discourse.