Support Us
Open Knowledge Foundation Open Knowledge
Foundation Blog

Why the German Digital Library should learn from Europeana

December 13, 2012 in Featured, Open GLAM

The full version of this article is available on the Open GLAM blog.

picture Launch of the DDB. Jill Cousins, Hermann Parzinger, Elke Harjes-Ecker, Matthias Harbort (from left to right) – Photo: Julia Hoppen
On the 29th of November 2012, the beta version of the German Digital Library (DDB) was officially launched. After five years of preparation and discussions with a large number of cultural institutions, it was finally time to bring it to the public. Herman Parzinger, head of the Prussian Cultural Heritage Foundation, explained in the press-release:

“The goal of the Deutsche Digitale Bibliothek (DDB) is to offer everyone unrestricted access to Germany’s cultural and scientific heritage, that is, access to millions of books, archived items, images, sculptures, pieces of music and other sound documents, as well as films and scores, from all over Germany”

To reach this goal, a lot of work needs to be done. At the moment, around 5.5 million metadata records can be found in the portal. Around 3 million come from a single institution, the Landesarchiv Baden-Württemberg. Currently 90 institutions provide data to the library and the three biggest organisations make up more than 80% of all records. Goal of the DBB is to include the metadata records of more than 30.000 German cultural institutions.

In many ways, the German Digital Library reminds of the Europeana project when it was launched in 2008. At that time, France was responsible for about 50% of all records in the Europeana portal and many countries were not present at all. In the past four years, Europeana has managed to include data from each EU country, and continues expanding it (see visualisation).

The interface of the DDB is very similar to Europeana as well. A simple search box combined with the possibility to filter the results in many different ways, for example by content provider, period, or location. As Europeana, the DDB is a search portal which links the user to the actual digitised object on the institutions webpage. They only host the metadata.

picture Homepage of the German Digital Library

Unfortunately, one major difference with the current Europeana project is how the DDB deals with copyright. Europeana has recently released all of their metadata records under a CC0 public domain waiver, making all of their metadata records free to use and reuse by anybody for any purpose without any restrictions.

The German Digital Library is quite the opposite. Their Terms of Use state clearly that:

  1. The DDB and its data suppliers retain all copyright and other industrial property rights to the public, freely accessible (and free of charge) digital content, derivatives and metadata including layout, software and their content in line with the principle of „free access – rights reserved“.
  2. Any use of the digital content, derivatives and metadata for commercial purposes is hereby prohibited, unless approved by the title holder in individual cases.

These copyright restrictions make it very hard for users to do anything with the metadata from the DDB. Especially when the API is launched, it is practically impossible for developers to create something with it as they will constantly have to ask the hundreds of different institutions if it is allowed. When Europeana started, there was also no consensus how to deal with the rights of the aggregated metadata and it took them four years to solve this issue. Over the last couple of years, the European Union, Europeana itself, and many other organisations have released reports and documents that clearly outline the advantages of open data for cultural institutions, as well as for society and research.

It seems like a strange move that the DDB is so restrictive, especially as they are to become the official German aggregator to Europeana. Europeana has been very clear since last September that the rights of all the data provided have to be waived away by using the CC0 declaration. Furthermore, many objects from for example the Landesarchiv Baden-Württemburg can already be found on Europeana, under a free license.

With all of the world’s heritage is becoming available online, great new possibilities arise. Different collections can be connected and linked and institutions can enrich their own data with the use of others. The history of Germany can not only be found in German institutions, but all over the world. By combining these different collections, historians can create a much more sophisticated history and find new stories and insights. This can only be achieved if the licenses being used by the different institutions and aggregators allow this, and the DDB term of use clearly do not do this.

As the German Digital Library is still in a beta-version, much can change. They are a direct partner of Europeana so it seems very easy to learn from the experiences of Europeana and how decisions in the past about copyright have worked out for them. Europeana has shown that European institutions are willing to provide data that can be freely reused, why start the discussion all over again in Germany?

Related posts:

  1. The Digital Public Library of America moving forward A fuller version of this post is available on the Open GLAM blog The Digital Public Library of America (DPLA) is an ambitious project to build a national digital library platform for the United States that will make the cultural...
  2. Europeana and Linked Open Data Europeana has recently released a new version of its Linked Data Pilot, data.europeana.eu. We now publish data for 2.4 million objects under an open metadata licence: CC0, the Creative Commons Public Domain Dedication. This post elaborates on this earlier one...
  3. 4 Stars for Metadata: an Open Ranking System for Library, Archive, and Museum Collection Metadata This post was written by participants of the LOD-LAM Summit which was held on June 2nd/3rd in San Francisco and is crossposted on the Creative Commons blog and the Open bibliography and Open Bibliographic Data blog. For author information see...

  • elleneuler

    Dear Joris,

    Thank you for your post, you are adressing a lot of important points! That said, we feel that the difference between the rights to metadata and the rights to digital objects (content and derivatives) themselves have not been sufficiently taken into account. You are discussing the metadata, yet the cited Terms of Use (TOS) focus on digital content and derivatives.

    The TOS of the DDB are merely an indication that there may be some rights reserved for metadata and digital objects which can be accessed through the portal, which must be observed. The TOS refer to the priority of explicitly defined rights and specific license terms and make it clear that where they are not named; the framework defined by laws such as the Copyright Law applies.

    Exactly the same applies, even if not explicitly described, for digital content and derivatives made accessible through Europeana. As for the metadata, there is a difference in as far as Europeana accepts only CC0 metadata, so only metadata that fulfils this requirement can be used to lead the user to digital content and derivatives and made available via an API.

    The API for metadata, which the DDB will provide in an initial version in the first half of 2013, will also only open metadata, which was provided by the cooperating partners under the CC0 license. However, this applies to the predominant share of our metadata.

    On the other hand, the museums in particular are not keen to make their metadata available under CC0, simply because their metadata is so elaborated and profound that – just like abstracts – it is eligible for copyright protection. They are not prepared to waive this protection completely. The motive behind this is less about trying to prevent commercial subsequent uses, but much more to keep some control over the data in the interest of safeguarding its authenticity and integrity in the further communication, so this is about a CC BY licence.

    To get an idea of the depth of the descriptive information in these cases, please just have a look at the digital objects of the Antikensammlung in the DDB, for example.

    Regarding the discussion with the museums, Europeana has the same issues as we do. But they chose to deal with out all metadata t is not under the CCO license. The DDB is not so strict in this regard, which allows us to benefit from the more profound metadata. This data can however only be passed on Europeana as defined core set of metadata (see the technical questions in our FAQ: http://www.deutsche-digitale-bibliothek.de/content/faq/#E) We will also not be able to make them available via the API.

    So, to sum it up: The DDB has made a compromise that allows for a elaborate metadata on one hand, and maximum reusability and openness of the metadata for the general public on the other.

    There is still a long way to go, but we are looking forward to the discussion.

    All the best,

    Dr. Ellen Euler Branch Office Manager Deutsche Digitale Bibliothek

    • Joris Pekel

      Dear Ellen,

      Thank you for your thorough response.

      I understand the issue about ‘when metadata becomes actual content’. Institutions who put a lot of effort in for example transcribing and translating documents often do not feel comfortable to waive away all rights, and perhaps rightfully so. This kind of content can get more value when the source is properly attributed. It does however seems like such a pity that because some institutions are not keen on doing this, nobody can benefit from openly licensed metadata which anyone can reuse and link. I am seriously wondering how many of the around 2000 institutions that signed the agreement are actually opposing CC0. I would even see a role for the OKFN and Wikimedia Germany to talk to these institutions about the benefits of open data.

      As you mentioned, Europeana allows the cultural institutions to choose what part of their metadata they will provide to Europeana under the CC0 license, and what not. This automatically leads to less high quality metadata, but it is freely available for anyone to reuse. Would it not be an option for the DDB to say, we want this data from your institution under CC0: name, year, author etc +thumbnail, and the full descriptions for example under a CC-BY-SA license? A filter option would not be hard to implement and with the API it should be even easier to achieve. The CC0 material can then be immediately be aggregated by Europeana, leading to higher visibility for the institutions and the DDB

      Finally, I think that the DDB, and for that matter Europeana as well, should put more effort in educating, convincing, and helping institutions to release digitised works that are in the public domain, and put inflict new copyright on scans of the work as they ar direct reproductions.

      The OKFN and myself would be very willing to help with these issues so do get in touch.

      All the best,

      Joris

Get Updates