Volume 18, No. 6, 2021

An Enhanced Approach Of RDF Graph Data In-Memory Processing For Social Networks With Performance Analysis

M. Baby Nirmala , J. G. R. Sathiaseelan


Ever-growing linked open data proliferate in size and number giving rise to “Big data problems”. Consequently, there is a rapid increase inf semantic data and services. It is significant to handle and provide more solutions to very large semantic data graphs are very significant. As the MapReduce algorithms are iterative and less effective, there is a need to go for in-memory processing. Ain of this paper is providing an enhanced RDF graph data processing approach with improved performance includes representation of RDF as a Directed labeled Multigraph, convert that RDF Graph as property Graph to have modeling workarounds (intermediate nodes), Reification and Singleton Property. Build this initial graph in high- performance analytic system SPARK with Graph X library using SCALA language to provide higher-order functions. So that functions and data structures can be stored in distributed memory and use SPARQL for query processing. The objective is of this paper is to propose an enhanced RDF Graph in-memory processing method that quickens the accessing of RDF Graph data, improves response times, and reduces the execution of time of RDF Graph data processing of this Social Network data.

Pages: 2157-2169

Keywords: Large Linked Open Data; In-memory processing; RDF; MapReduce; Spark; Graph X; Scala.

Full Text