Expansion of a relational database to support semantic web queries

Abstract

Nowadays various types of data, especially those with public character, are stored and represented in relational databases. While this way of storing data is a practice for most of the institutions, it turns them into isolated silos with low level of accessibility and interoperability in the web. Indeed not all datasets are readable from the World Wide Web, therefore to increase their access it is necessary to provide mapping from relational databases to any serialization format of Resource Description Framework (RDF), as one of the best practices for distributing and interlinking data on the web. While there are several ways to accomplish the mapping process, the most appropriate and fastest way is by using the Protege Plugins application. In this paper we present this mapping method, and describe the complete conversion of a database from SQL Server relational Database to RDF using different scenarios. By using Sesame Repository, as the local repository for our dataset, several SPARQL queries will be generated on top of our data. As a main contribution this paper emphasizes the advantages and performance of querying and using serialized RDF statements versus those stored in the relational databases.