Efficient SPARQL Evaluation on Stratified RDF Data with Meta-data

The Resource Description Framework (RDF) is a simple, but frequently used W3C standard, which uses triplets to define relationships between resources. In this paper the evaluation of queries in the query language SPARQL on RDF data with meta-data is investigated. We first show that if the data are stratified, i.e. a particular partial order can be defined on the meta-data labels, then a nesting procedure can be applied, which induces a rewriting of the query. Based on a specification by an Abstract State Machine we show that the result of the rewritten query equals the one that would have resulted from the evaluation of the original query. We further investigate the reduction of complexity by using data and query nesting.