momo54 / semLAV

SemLAV allows to execute SPARQL queries over Deep Web and Linked Open Data data sources. It follows the mediator approach based on view definitions over remote data sources. A query is expressed using a mediator schema, SemLAV selects relevant data sources and rank them. The ranking strategy is designed to deliver results quickly based only on view definitions, i.e., no statistics, nor probing on sources are required. In this demonstration, we validate the effectiveness of SemLAV approach with real data sources from social networks and Linked Open Data. We show in different setups that materializing only a subset of ranked relevant views is enough to deliver significant part of expected results. Ref: G. Montoya, L. D. Ibáñez, H. Skaf-Molli, P. Molli, and M.-E. Vidal. Semlav:Local-as-view mediation for sparql. Transactions on Large-Scale Data- and Knowledge-Centered Systems XIII, Lecture Notes in Computer Science, Vol. 8420, pages 33-58,2014. URL: http://tldks.faw.at/paper/92/