Design and evaluation of a NoSQL database for storing and querying RDF data

Abstract

Currently the amount of web data has increased excessively. Its metadata is widely used in order to fully exploit
web information resources.This causes the need for Semantic Web technology to quickly analyze such big data.
Resource Description Framework (RDF) is a standard for describing web resources. In this paper, we propose
a method to exploit a NoSQL database, specifically MongoDB, to store and query RDF data. We choose
MongoDB to represent a NoSQL database because it is one of the most popular high-performance NoSQL
databases. We evaluate the proposed design and implementation by using the Berlin SPARQL Benchmark,
which is one of the most widely accepted benchmarks for comparing the performance of RDF storage systems.
We compare three database systems, which are Apache Jena TDB (native RDF store), MySQL (relational
database), and our proposed system with MongoDB (NoSQL database). Based on the experimental results
analysis, our proposed system outperforms other database systems for most queries when the data set size is
small. However, for a larger data set, MongoDB performs well for queries with simple operators while
MySQL offers an efficient solution for complex queries. The result of this work can provide some guideline for
choosing an appropriate RDF database system and applying a NoSQL database in storing and querying RDF
data.