Scalable peer-to-peer-based RDF management

Abstract

Handling web-scale RDF data requires sophisticated data management that scales easily and integrates seamlessly into existing analysis workflows. We present Hdrs-- a scalable storage infrastructure that enables online-analysis of very large RDF data sets. Hdrs combines state-of-the-art data management techniques to organize triples in indexes that are sharded and stored in a peer-to-peer system. The store is open source and integrates well with Hadoop MapReduce or any other client application.