%0 Conference Paper %1 Urbani2010WebPIE %A Urbani, Jacopo %A Kotoulas, Spyros %A Maassen, Jason %A van Harmelen, Frank %A Bal, Henri %B Proceedings of the 7th Extended Semantic Web Conference (ESWC) %C Berlin, Heidelberg %D 2010 %E Aroyo, Lora %E Antoniou, Grigoris %E Hyvönen, Eero %E ten Teije, Annette %E Stuckenschmidt, Heiner %E Cabral, Liliana %E Tudorache, Tania %I Springer %K %P 213--227 %T OWL Reasoning with WebPIE: Calculating the Closure of 100 Billion Triples %V 6088 %X In previous work we have shown that the MapReduce frame- work for distributed computation can be deployed for highly scalable inference over RDF graphs under the RDF Schema semantics. Unfortu- nately, several key optimizations that enabled the scalable RDFS infer- ence do not generalize to the richer OWL semantics. In this paper we analyze these problems, and we propose solutions to overcome them. Our solutions allow distributed computation of the closure of an RDF graph under the OWL Horst semantics. We demonstrate the WebPIE inference engine, built on top of the Hadoop platform and deployed on a compute cluster of 64 machines. We have evaluated our approach using some real-world datasets (UniProt and LDSR, about 0.9-1.5 billion triples) and a synthetic benchmark (LUBM, up to 100 billion triples). Results show that our implementation is scal- able and vastly outperforms current systems when comparing supported language expressivity, maximum data size and inference speed.

@inproceedings{Urbani2010WebPIE, abstract = {In previous work we have shown that the MapReduce frame- work for distributed computation can be deployed for highly scalable inference over RDF graphs under the RDF Schema semantics. Unfortu- nately, several key optimizations that enabled the scalable RDFS infer- ence do not generalize to the richer OWL semantics. In this paper we analyze these problems, and we propose solutions to overcome them. Our solutions allow distributed computation of the closure of an RDF graph under the OWL Horst semantics. We demonstrate the WebPIE inference engine, built on top of the Hadoop platform and deployed on a compute cluster of 64 machines. We have evaluated our approach using some real-world datasets (UniProt and LDSR, about 0.9-1.5 billion triples) and a synthetic benchmark (LUBM, up to 100 billion triples). Results show that our implementation is scal- able and vastly outperforms current systems when comparing supported language expressivity, maximum data size and inference speed.}, added-at = {2011-12-12T19:00:44.000+0100}, address = {Berlin, Heidelberg}, author = {Urbani, Jacopo and Kotoulas, Spyros and Maassen, Jason and van Harmelen, Frank and Bal, Henri}, biburl = {https://www.bibsonomy.org/bibtex/2228a3c656aed1d3e26a239fe957b6903/gergie}, booktitle = {Proceedings of the 7th Extended Semantic Web Conference (ESWC)}, editor = {Aroyo, Lora and Antoniou, Grigoris and Hyvönen, Eero and ten Teije, Annette and Stuckenschmidt, Heiner and Cabral, Liliana and Tudorache, Tania}, file = {:Urbani2010WebPIE.pdf:PDF}, groups = {public}, interhash = {a9c94b69cfa9ebbf8f44c0ff6ac23190}, intrahash = {228a3c656aed1d3e26a239fe957b6903}, keywords = {}, month = May, pages = {213--227}, publisher = {Springer}, series = {Lecture Notes in Computer Science}, timestamp = {2011-12-12T19:00:44.000+0100}, title = {{OWL Reasoning with WebPIE: Calculating the Closure of 100 Billion Triples}}, username = {gergie}, volume = 6088, year = 2010 }