One year later: Linked Open Data in the German National Library

A little more than a year ago, the German National Library (DNB) announced that it would release more data as linked data under an open license. It was decided that the metadata would be released with as little restrictions as possible by using the CC0 rights waiver. This means that anybody can use and reuse the data in any way possible, also for commercial purposes.

Now one year later, we talk with Lars G. Svensson, Advisor for Knowledge Networking at the DNB, about what this move has meant for the library.

Welcome Lars, thank you for taking the time.

Thank you for having me!

Could you tell me why the library decided to open up the metadata?

In September 2011 the Conference of European National Librarians (CENL) decided to adopt CC0 licensing for their data. The DNB had started to publish authority data as linked data in spring 2010. We first used a home-grown license based partly on CC BY-SA but with the restriction that commercial entities needed to register before they can use the data. Since our Director General had been one of the supporters of the CENL decision it was natural for us to move in the same direction. One of the key points with linked data is that other people have to be able to reuse and connect the data with other sources. For that reason we decided last year to discontinue the license based on CC BY-SA and go for CC0 in order to have as few restrictions as possible for reuse. Currently, we publish two datasets: The first one is the authority data, which consists of data for names of persons, organizations, events, places, and works. Since January 2012 there is also bibliographic data available with title, publisher etc., which re-uses the authority data. The data is available under CC0 in many formats including RDF. The only exception is bibliographic data in library specific formats (MARC 21 and MARC XML) from the last two years but we expect that this restriction will disappear after 2015.

And have you seen interesting cases of reuse so far?

Yes definitely. One of my favourite projects is the Museum Digital. This is a German digital open platform where smaller institutions can put their content. The museums curate and manage their own database on the site and enter their own metadata. The site included our metadata to create more links from and to the content available on the platform. They also found out that we include a link to DBpedia in our data. This allowed them to import that data into the platform in various languages. This greatly enriches the information on the platform.

Not all libraries are in the position to release their own metadata because they make use of services and are therefore not the owners of the data. How does that work in the DNB?

We are in the fortunate position to be the national library, so it is basically our job to create this data in the first place. That allows us to freely distribute it in any way we want to. The authority data is curated together with the German library networks, so that is not really our data, but it was not a problem to agree on the open license.. As we are all public institutions, openness helps us to reach out to the public.

Does the German National Library also provide access to digitised books?

We are a relatively young library which was founded in 1913. For that reason we don’t have that much material that is in the public domain. So we do digitise our collection, but since we are not the owners of the rights we can only show the material to people in the reading rooms in the library. We try to make the books that are out of copyright as accessible as possible. We started for example with a collection with 100 classic books such as the works of Goethe and Schiller. These are freely accessible – also in Europeana – as they are in the public domain and we currently have large digitisiation projects also comprising out-of-copyright material. A further service we offer is digitization of tables of contents; Those are very popular among our users since they offer both more terminology we can index in our catalogue and more contextual information making it easier for our patrons to decide whether the publication they found suits their needs or not.

Great to hear, and what’s next for the library?

We are still in the transition phase so not all metadata is yet openly available in all formats, we expect that this will happen in the next few years and then our metadata will be completely open. We keep improving our linked datasets and work hard to also get to make more content available.

That’s great, thank you very much for taking the time for this interview!

My pleasure!