A Benchmark Suite for Federated SPARQL Query Processing from Existing Workflows

Tracking #: 1594-2806

This paper is currently under review
Antonis Troumpoukis
Angelos Charalambidis
Giannis Mouchakis
Stasinos Konstantopoulos
Daniela Digles
Ronald Siebes
Victor de Boer
Stian Soiland-Reyes

Responsible editor: 
Guest Editors Benchmarking Linked Data 2017

Submission type: 
Full Paper
This paper presents a new benchmark suite for SPARQL query processors. The benchmark is derived from workflows established by the pharmacology community and exploits the fact that these workflows are not only applied to voluminous data, but they are also equivalent to complex and challenging queries. The value of this queryset is that it realistically represents actual community needs in a challenging domain, testing not only speed and robustness to large data volumes but also all features of modern query processing systems. In addition, the natural partitioning of the data into meaningful datasets makes these workflows ideal for benchmarking federated query processors. This emphasis on federated query processing drived complementing the benchmark with an execution engine that can reproduce distributed and federated query processing experiments.
Full PDF Version: 
Under Review