Date of this Version
Linked open data (LOD) is an essential component in semantic web architecture and is becoming increasingly important over time due to its ability to share and re-use structured data which is both human and computer readable over the web. Currently, many libraries, archives, museums etc. are using open source digital library software to manage and preserve their digital collections. They may also intend to publish their e-resources as “Linked Open Datasets” for further usage. LOD enables the libraries or information centers to publish and share the structured metadata that is generated and maintained with their own bibliographic and authority data in such a way that the other libraries and general community across the world can consume, interact, enrich and share. In this context, the key issue is to convert the library bibliographic data which is commonly known as metadata into LOD dataset. The purpose of this paper is to provide a methodology and technical aspects to design and publish a structured LOD dataset of bibliographic information from a digital repository developed with DSpace digital library software so that other libraries can link their repositories with these LOD for providing additional relevant resources to their end-users. The paper shows the process of integration and configuration of Apache Jena Fuseki (a tool for SPARQL Endpoint interface) with DSpace for converting metadata into Resource Description Framework (RDF) triple model and make them available in various RDF formats. It also discusses a model for building a LOD framework to convert and store RDF graph and RDF triple. Finally, it tests the accessibility of the inked open dataset by querying RDF data through a SPARQL endpoint interface.