RDFGrowth, a P2P annotation exchange algorithm for scalable semantic web applications
2004 (English)In: CEUR Workshop Proceedings, ISSN 1613-0073, Vol. 108Article in journal (Refereed) Published
We present RDFGrowth, an algorithm that addresses a specific yet important scenario: large scale, end user targeted, metadata exchange P2P applications. In this scenario, peers perform browsing and querying of semantic web statements on a local database without directly generating network traffic or remote query execution. The database grows by learning from other peers in the P2P group using only a minimal amount of direct queries that are guaranteed to be executable with a low, predictable computational cost. Although full RDF graphs could be treated, the design allows a peer to learn only about resources considered interesting by a specific "community" and makes it possible to tag the received information according to individual trust rules. Inspired by well known viral distributed information techniques, the algorithm is in agreement with the RDF semantics and is specifically suited for the properties of Distributed Hash Table P2P networks. A few assessments about the applicability of RDFGrowth to real semantic web applications are also given.
Place, publisher, year, edition, pages
2004. Vol. 108
Algorithms; Distributed computer systems; Knowledge management; Query languages; Query processing; Semantic Web; Social networking (online); World Wide Web, Computational costs, Distributed Hash Table, Distributed information, Exchange algorithms, Network traffic, P2P applications, Query execution, Semantic web applications, Peer to peer networks
IdentifiersURN: urn:nbn:se:kth:diva-174756ScopusID: 2-s2.0-84930701931OAI: oai:DiVA.org:kth-174756DiVA: diva2:859625
QC 201510082015-10-082015-10-072015-10-08Bibliographically approved