Dynamic Join Order Optimization for SPARQL Endpoint Federation

The existing web of linked data inherently has distributed data sources. A federated SPARQL query system, which queries RDF data via multiple SPARQL endpoints, is expected to process queries on the basis of these distributed data sources. During a federated query, each data source may consist of a search space of nontrivial size. Therefore, finding the optimal join order to minimize the size of intermediate results from different sources is key to optimizing the performance of such federated queries. In this study, we present a dynamic optimization approach to determining join order, which can find more optimized join plans than static optimization approaches. Our experimental results show that our proposed approach stably improves the performance of a federated query as the query becomes increasingly complex.