Towards a reference architecture for Semantic Web applications

Abstract

The Semantic Web currently has two complementary architectural approaches: ¿Bottom-up¿ emergent best practices by the community and ¿top-down¿ prescriptive standards by standards bodies, leaving a gap regarding the concrete implementation of Semantic Web technologies. Based on the Web Science approach of combining empirical analysis with engineering, we are proposing to fill this gap with a reference architecture consisting of seven reusable and domain independent components, extracted from applications submitted at key demonstration challenges in the Semantic Web domain. The reference architecture can be used (1) as a common terminology for communicating concepts and issues related to the implementation of Semantic Web technologies, and (2) as a blueprint when implementing SemanticWeb applications.