5.7. Experimental analysis

In this section, we report on our extensive experiments on two RDF data sets. We evaluate the time performance and the robustness of AMBER w.r.t., the state-of-the-art competitors by varying the size and structure of the SPARQL queries. Experiments are carried out on a 64-bit Intel Core i7-4900MQ @ 2.80GHz, with 32GB memory, running Linux OS - Ubuntu 14.04 LTS. AMBER is implemented in C++.

5.7.1. Experimental setup

We compare AMBER with the four standard RDF engines: Virtuoso-7.1 [ERL 12], x-RDF-3X [NEU 10], Apache Jena [CAR 04] and gStore [ZOU 14b]. For all these competitors, we use the source code available on the website or obtained by the authors. Another recent work, Turbo_HOM++ [KIM 15], has been excluded, since it is not publicly available.

For experimental analysis, we use two RDF data sets: DBPEDIA and YAGO. DBPEDIA constitutes the most important knowledge base for the Semantic Web community. Most of the data available in this data set come from the...