Large Scale RDF Storage Solutions Evaluation

Abstract

The increasing popularity of the SemanticWeb and Semantic Technologies require sophisticated ways to store huge amounts of semantic data. RDF together with the rule base RDF Schema have proved themselves as good candidates for storing semantic data due to the simplicity and high abstraction level. A number of large scale RDF data storage solutions have been proposed. Several typical representative have been discussed and compared in this work, namely Sesame, Kowari, YARS, Redland and Oracle's RDF MATCH table function. We present a comparison of those approaches with respect to consideration of context information, supported access protocols, query languages, indexing methods, RDF Schema awareness, and implementation. We also identify applicability as well as discuss advantages and disadvantages of particular approach. Furthermore, an overview of storage requirements and performance tests has been presented. A summary of performance analysis and recommendations are given and discussed.