Federated Query Processing for the Semantic Web

During the last years, the amount of RDF data has increased exponentially over the Web, exposed via SPARQL endpoints. These SPARQL endpoints allow users to direct SPARQL queries to the RDF data. Federated SPARQL query processing allows to query several of these RDF databases as if they were a single one, integrating the results from all of them. This is a key concept in the Web of Data and it is also a hot topic in the community. Besides that, the W3C SPARQL-WG has standardized it in the new Recommendation SPARQL 1. 1. This book provides a formalization of the W3C proposed recommendation. This formalization allows to identify existing errors and allows to correct them before the implementation phase, or when the execution of these federated queries start. The book constitutes a valuable resource for any implementer since it also proposes solutions to the problems identified as well as proposing a set of SPARQL pattern reordering rules, which reduce the execution time of federated queries significantly. Another strong point of this book is the research methodology followed. It states clearly the problems in the state of the art, next defines the research hypothesis for next providing a thoroughly analysis of the semantics of the SPARQL 1. 1 specification. Once the theoretical part is concluded, the book steps into the implementation part, describing clearly the implementation decisions for finally evaluating the overall system.