Home // DBKDA 2014, The Sixth International Conference on Advances in Databases, Knowledge, and Data Applications // View article

Design of Distributed Storage Manager for Large-Scale RDF Graphs

Authors:
Iztok Savnik
Kiyoshi Nitta

Keywords: databases; RDF databases; distributed database systems; query processing system; database system implementation

Abstract:
Storage and management of large-scale RDF repositories is a challange that may be compared to storage and management of large-scale HTML repositories. The main difference is in the modelling power of RDF data model comparing it to HTML hyper-graph model. RDF is much closer to the database data model and requires the capabilities of database management system rather than those offered by informational retrieval query engine. Triple-based storage systems seem to provide the functionality needed for storing RDF graphs. Triples are very natural means for the representation of graphs at various levels of abstraction. To cope with growing demand for querying large-scale triple-stores including up to several Tera triples we propose the use of massively parallel system that can be dynamically configured for particular queries into a set of parallel data-flow machines. The paper presents the design of large-scale triple-store database system.

Pages: 154 to 160

Copyright: Copyright (c) IARIA, 2014

Publication date: April 20, 2014

Published in: conference

ISSN: 2308-4332

ISBN: 978-1-61208-334-6

Location: Chamonix, France

Dates: from April 20, 2014 to April 24, 2014