Peer-based query rewriting in SPARQL for semantic integration of linked data

In this proposal we address the problem of ontology-based SPARQL query answering over distributed Linked Data sources, where the ontology is given by conjunctive mappings between the source schemas in a peer-to-peer fashion and by equality constraints between constants. In our setting, the data is not materialised in a single datastore: it is accessed in a distributed environment through SPARQL endpoints. We aim to achieve query answering by generating the perfect rewriting of the original query and then processing the rewritten query over distributed SPARQL endpoints. We identify a subset of ontology constraints that enjoy the first-order rewritability property and we perform preliminary empirical evaluation taking into account such restricted constraints only. For future work, we aim to tackle the query answering problem in the general case