Exposing scholarly information as Linked Open Data: RDFizing DSpace contents

Abstract

This paper introduces a transformation engine which can be used in order to convert an existing institutional repository installation into a Linked Open Data repository. We describe how the data that exists in a DSpace repository can be semantically annotated in order to serve as a Semantic Web (meta)data repository. We present a non-intrusive, standardscompliant approach that can run alongside with current practices, while incorporating stateof-the art methodologies. Also, we propose a set of mappings between domain vocabularies that can be (re)used towards this goal.

Exposing Bibliographic Information as Linked Open Data Using Standards-based Mappings: Methodology and Results

“…Based on a mapping file for each repository, and using the tool introduced in (Konstantinou, Spanos, Houssos, & Mitrou, 2014), we were able to extract the database contents in an RDF graph. An example of how the mappings are formed, using R2RML, follows next.…”

“…These include the authors' previous work, introduced in (Konstantinou, Spanos, Houssos, & Mitrou, 2014), an approach that appears to perform much faster in RDF-izing relational database contents compared to real-time SPARQL-to-SQL translators. This option is being supported by D2RQ as well alongside its main function as an RDF server.…”

Materializing the Web of Linked Data

“…Its successor, Morph (Priyatna et al 2014 ) supports R2RML mappings, offering SPARQL-based data access to the contents of a database. (Konstantinou et al 2014a ), (Konstantinou et al 2013a ) is an open-source tool that can export RDF graphs from the contents of a relational database, based on an input R2RML mapping fi le. Its main components are a parsing engine, which analyzes an R2RML mapping fi le and materializes the specifi ed RDF graph and a faceted browser of the generated RDF graph.…”

“…5 . Thus, although the asynchronous export approach from a relational database to RDF could be followed in many domains, it fi nds ideal application in the scholarly literature domain and, in general, in cases when data is not updated to a signifi cant amount daily (Konstantinou et al 2014a). In these cases, when freshness is not crucial and/or selection queries over the contents are more frequent than the updates, what our approach succeeds in, is a lower downtime of the system that hosts the generated RDF graph and serves query-answering.…”

“…Overall high-level architectural overview and information fl ow in the approach(Konstantinou et al 2014a ) …”