FMQO: A Federated RDF System Supporting Multi-query Optimization

This demo designs and implements a system called FMQO that can support multiple query optimization in federated RDF systems. Given a set of queries posed simultaneously, we propose a heuristic query rewriting-based approach to share the common computation during evaluation of multiple queries. Furthermore, we propose an efficient method to use the interconnection topology between SPARQL endpoints to filter out irrelevant sources and join intermediate results during multiple query evaluation. The experimental studies over both real federated RDF datasets show that the demo is effective, efficient and scalable.