RDF keyword search query processing via tensor calculus

Abstract

Keyword-based search over (semi)structured data is considered today an essential feature of modern information management systems and has become an hot area in database research and development. Answers to queries are generally sub-structures of a graph, containing one or more keywords. While finding the nodes matching keywords is relatively easy, determining the connections between such nodes is a complex problem requiring on-the-fly time consuming graph exploration. Current techniques suffer from poorly performing worst case scenario or from indexing schemes that provide little support to the discovery of connections between nodes. In this paper we propose an indexing scheme for RDF graphs based on the first principles of linear algebra, in particular on tensorial calculus. Leveraging our abstract algebraic framework, our technique allows to expedite the retrieval of the sub-structures representing the query results. © 2012 Springer-Verlag.