Approximating Inference-enabled Federated SPARQL Queries on Multiple Endpoints

Abstract. Running inference-enabled SPARQL queries may sometimes require unexpectedly long execution time. Therefore, demand has in-creased to make them more usable by slightly changing their queries, which could produce an acceptable level of similar results. In this demon-stration, we present our query-approximation system that can transform an inference-enabled federated SPARQL query into another one that can produce acceptably similar results without unexpectedly long runtimes to avoid timeout on executing inference-enabled federated SPARQL queries