Rewriting queries on SPARQL views

摘要: The problem of answering SPARQL queries over virtual views is commonly encountered in a number settings, including while enforcing security policies to access RDF data, or when integrating data from disparate sources. We approach this by rewriting the equivalent underlying thus avoiding costs entailed view materialization and maintenance. show that query combines most challenging aspects for relational XML cases: like case, requires synthesizing multiple views; size rewritten exponential views. In paper, we present first native algorithm SPARQL. For an input set views, resembles union conjunctive can be size. propose optimizations basic (i) minimize each union; (ii) eliminate with empty results evaluation; (iii) efficiently prune out big portions search space rewritings. experiments, performed on two stores, our algorithms are scalable independent stores. Furthermore, have order magnitude improvements both evaluation time.