3rdf: Storing and Querying RDF Data on Top of the 3nuts Overlay Network

In current research Peer-to-Peer (p2p) based Semantic Web systems mainly use distributed hash table (DHT) based networks. These networks provide good load balancing by applying uniform hash functions with the drawback that they destroy possible semantic relations between data elements. But mapping the data semantics on the network structure could improve the routing time in the network and consequently the RDF query latency on application layer. In this paper, we present 3rdf, a distributed RDF system for storing and querying RDF data. The 3rdf system has been built on top of the 3nuts p2p network. The 3nuts network improves on reducing the query response time and bandwidth usage in our system by adapting the network structure to the semantics of the RDF data. In addition, we study how the evaluation of SPARQL BASIC graph patterns in existing distributed RDF repositories can be extended for other graph patterns such as OPTIONAL and UNION in our 3rdf system.