An interactive session on Linked Data - 3 talks:

1. The 99 cents Linked Open Data archive - Miel Vander Sande

If a hosting a queryable Linked Open Data endpoint is a bit pricey, hosting a whole archive is really expensive. Not from our perspective. In this talk, we present a sustainable strategy to publish LOD with a small budget and still allow users to query it at will. In addition, we show how historical queries can be easily supported on prior dataset versions. As a use case, we refurbished the DBpedia archive and demonstrate its function live.

Joinup (joinup.ec.europa.eu) aggregates open source software and standards produced by european public administrations. At it's core it uses ADMS-AP, a DCAT extension as it's data model. In the upcoming release of Joinup, we opted for an architecture where a triplestore is the cornerstone of the application, and a generic SPARQL connector has been build.This allows editors to create RDF datasets, without being aware of RDF altogether.

The Belgian War Press contains a lot of interesting newspapers about the First World War. They have been digitised and made searchable with OCR. A user can only find information by trial and error in user-friendly search interfaces. This prevents machines to resolve more complex queries. In this session we’ll cover how Linked Open Data can help us to solve this problem and how to clean and publish this data on the Web.

Feel free to share your notes with us!