Matthias Palmér from MetaSolutions gave two presentations in Oslo on November 6, by invitation of the Directorate for Cultural Heritage at Riksantikvaren. Roughly 35 persons with an interest in Linked Data for Cultural Heritage attended.

The first talk focused on how to take steps towards making the service Norvegiana more LOD-friendly. The presentation is available on Slideshare, although in Swedish:

The take home message was basically that if you are to adapt to the Europeana Data Model (EDM) it is beneficial to separate metadata about things, their digital representations and the metadata records that describe them. This follows in large from the documentation on EDM, nevertheless considering the consequences was worthwhile. For instance, should Norvegiana do the aggregation of many sources or should that be the responsibility of Europeana? And if it is left to Europeana, what purpose does Norvegiana have, other than an intermediate harvester provided for convenience? And what happens with portals or apps that have a focus on Norway, should those use Europeana rather than Norvegiana?

Matthias also provided some advice on practicalities on how to model information in RDF.

The second presentation had a complementary focus on how a LOD-friendly Norvegiana can be used by data consumers such as the mobile app developed within the project Kultur og naturreise, which includes a mobile app with a focus on providing information about culture and nature. Also this presentation is in Swedish and published at Slideshare:

In the presentation Matthias argued that the benefits of using the LOD version of Norvegiana can be divided into three areas. First it would provide fewer but more information dense information objects on a map. Second, it would allow links to be followed between information objects without leaving the app. This has the additional benefit of also allowing information objects that do not have a position in the map to be discovered. Third, it can help to provide recommendations for other information objects based on what is being viewed at the moment.