Classification of Knowledge Processing by MapReduce

The representation of a semantic data on web through RDF (Resource Description Framework) is in strong increase these last years. Treatment of data sets RDF by the traditional solutions of the centralised systems generates various problems in data access and its recovery. To resolve it, several solutions were recently proposed to manage big quantities of semantic data in distributed systems with the use of MapReduce framework for treatment, management, and stocking of these data. The purpose of this paper is to present the main solutions proposed in this context. A typology of the different methods is based on the use of MapReduce, Pig's functions, partitioning RDF data, storing data, and query processing.