The Semantic Web vision promises an extension of the current Web in which all data is annotated with machine understandable metadata. The relationship-centric nature of this data has led to the definition of Semantic Associations, which are complex relationships between resources. Semantic Associations attempt to answer queries of the form “how are resource A and resource B related?” Knowing how two entities are related is a crucial question in knowledge discovery applications. Much the same way humans collaborate and interact to form new knowledge, discovery of Semantic Associations across repositories on a peer-to-peer network can allow peers to share their local knowledge to collectively make new discoveries. In this paper we propose a method for computing Semantic Associations over distributed RDF data stores in a peer-to-peer setting. We follow a hierarchical peer / super-peer network topology, and we propose a novel query planning algorithm based on a notion of knowledgebase borders and minimum distances between borders.
Arpinar, I. B.,
& Sheth, A. P.
(2005). Peer-to-Peer Discovery of Semantic Associations. .
Additional FilesP2P.pdf (437 kB)