A Semantic Matching Approach for Distributed RDF Data Query on a Knowledge Bus

Download

PDF - Published Version Download (199Kb)

Description/Abstract

During the past several years, semantic web technologies have been applied to facilitate the sharing of data in new and unexpected ways. Various heterogeneous data are assigned well-defined meaning by ontologies and expressed with RDF triples so that they are are able to be integrated and extracted across multiple sources. As the order of magnitude of triples hosted is higher and higher, and considering other essential issues (e.g. copyright, security) in application domains, a distributed RDF data management approach is more appropriate for knowledge sharing and integration. In this paper, we present a knowledge bus infrastructure - a general solution of locating and extracting knowledge elements from distributed sources on-demand rather than loading all of RDF triples into a large central triple store in advance. A semantic matching approach is discussed to support the key function of automatic knowledge source location. The knowledge bus infrastructure as well as the semantic matching approach has been adopted in the CFMS project, enabling a rapid information search and access for the engineering domain.