The German National Library (DNB) has launched a Linked Data version of the German National Bibliography.
The bibliographic data of the DNB’s main collection (apart from the printed music and the collection of the Deutsches Exilarchiv) and the serials (magazines, newspapers and series of the German Union Catalogue of serials (ZDB)) have been converted. Henceforth the RDF/XML-representation of the records are available at the DNB portal. This is an experimental service that will be continually expanded and improved.
This is a welcome extension to their Linked Data Service, previously delivering authority data. Documentation on their data and modelling is available, however the English version has yet to be updated to reflect this latest release.
Links to RDF-XML versions of individual records are available directly from the portal user interface, with the usual Linked Data content negotiation techniques available to obtain HTML or RDF-XML as required.
This is a welcome addition to the landscape of linked open bibliographic data, joining others such as the British Library.
Also to be welcomed is their move to CC0 licensing removing barriers, real or assumed, to the reuse of this data.
I predict that this will be the first of many more such announcements this year from national and other large libraries opening up their metadata resources as Linked Open Data. The next challenge will be to identify the synergies between these individual approaches to modelling bibliographic data and balance the often competing needs of the libraries themselves and potential consumers of their data who very often do not speak ‘library’.
Somehow [without engaging in the traditional global library cooperation treacle-like processes that take a decade to publish a document] we need to draw together a consistent approach to modelling and publishing Linked Open Bibliographic Data for the benefit of all – not just the libraries. With input from the DNB, British Library, Library of Congress, European National Libraries, Stanford, and others such as Schema.org, W3C, Open Knowledge Foundation etc., we could possibly get a consensus on an initial approach. Aiming for a standard would be both too restrictive, and based on experience, too large a bite of the elephant at this early stage.