Spar(k)ql: SPARQL Evaluation Method on Spark GraphX

Abstract

RDF is a flexible data representation model. Due to its flexibility and simplicity it has become a popular framework, hence the amount of RDF data is increasing fast. Querying massive amount of data is a serious challenge in general and it is true for RDF data as well. In this paper we investigating how to evaluate SPARQL queries on a distributed system. We propose a novel method for evaluating SPARQL queries using the GraphX graph analytical tool. GraphX is built on the top of Spark that is an in-memory data processing system for distributive computation. With this tool we managed to utilize the graph-like structure of the RDF statements which are in form of subject-predicate-object.

DOI: 10.1109/W-FiCloud.2016.48

7 Figures and Tables

Cite this paper

@article{Gombos2016SparkqlSE, title={Spar(k)ql: SPARQL Evaluation Method on Spark GraphX}, author={Gergo Gombos and G{\'a}bor R{\'a}cz and Attila Kiss}, journal={2016 IEEE 4th International Conference on Future Internet of Things and Cloud Workshops (FiCloudW)}, year={2016}, pages={188-193} }