Index-Based Source Selection and Optimization
Jürgen Umbrich, Marcel Karnstedt, Axel Polleres, Kai-Uwe Sattler · 2015
Digital Enterprise Research Institute, National University of Ireland, Galway Marcel Karnstedt Digital Enterprise Research Institute, National University of Ireland, Galway Axel Polleres Siemens AG Österreich, Austria Kai-Uwe Sattler Ilmenau University of Technology, Germany The recent developments around Linked Data have led to the exposure of large amounts of data on the Semantic Web amenable to automated processing in software programs [99]. Linked Data sources use RDF (the Resource Description Framework) in various serialization syntaxes for encoding graphstructured data, and the Hypertext Transfer Protocol (HTTP) for data transfer. Currently only a few data sources on the Web provide full query processing capabilities (e.g., by implementing SPARQL [142,444], a query language and protocol for RDF); there are millions of RDF documents but only around 360 public SPARQL endpoints.