Using semantic web technologies in the DCA project

One of the core work packages in the DCA project is investigating metadata and their role in using semantic web technologies in relation to the content that will be digitised within the project. At the end of 2012, work package leader iMinds released two new deliverables around this topic.

D3.2 Recommendations on Contextualisation and Enrichment of Contemporary Art
Authors: Sam Coppens & Erik Mannens, iMinds

This deliverable elaborates on applying semantic web technologies in order to contextualise and enrich your information.

First, some basic semantic web building blocks are introduced:
- data descriptions using RDF and URIs;
- adding semantics using RDFS and OWL;
- querying RDF using SPARQL;
- publishing RDF as Linked Open Data.

After this introduction, some best practices for describing and contextualising data are given. Contextualisation takes place on several levels. It starts with the data model and ensures the achievement of some interoperability with other data models or even the inheritance of some of their semantics. Then, some best practices for the data descriptions are introduced. The idea here is that the data descriptions must have enough information to describe the data unambiguously. If one is able to achieve this, a good context is created. Finally, this is followed by a discussion on how to enrich data with external data sources, to maximise the data contextualisation.

One way of contextualising data is through the use of controlled vocabularies. Special attention goes to the introduction of SKOS for describing controlled vocabularies. Also here, some best practices on how to model vocabularies are given. In the last chapter, these best practices are applied to the DCA vocabulary, designed to support multilingual search in Europeana.

Appendix: D5.3 Enrichment module and POC
Authors: Sam Coppens & Erik Mannens, iMinds

In this deliverable, we discuss how to publish data as Linked Open Data in practice. This means we show how to do it, which steps to take and which open source tools can be used for the different steps.

The basic steps in publishing data as Linked Open Data are:
- select appropriate RDF model to publish the data;
- choose a Linked Open Data server infrastructure;
- transform the data to RDF;
- enrich the data.

Each of the steps is discussed in detail and demonstrated by our proof of concept. An important step in publishing Linked Open Data is to enrich the data. In this deliverable, special attention will be given to the enrichment module. Finally, we’ll present this deliverable with a workflow recommendation for the DCA partners to publish their data as Linked Open Data.

The deliverable D3.2 Recommendations on Contextualisation and Enrichment of Contemporary Art and its Annex D5.3 Enrichment module and POC can be retrieved from the public deliverables section.

Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License /  Privacy /