Linked Open Data Bibliographic data on the Internet

Adrian Pohl
Adrian Pohl | Photo (detail): private

Adrian Pohl outlines the opportunities which arise when bibliographic data are made openly and freely available. At the central university library of the state of North Rhine-Westphalia, he is responsible for “Open Data and Linked Data” and coordinates the Working Group on Open Bibliographic Data of the Open Knowledge Foundation.

Mr Pohl, you are the coordinator of the Working Group on Open Bibliographic Data of the Open Knowledge Foundation. What is the remit of this international working group?

The group has set itself the task of giving the public free and unrestricted access to bibliographic data. These data describe books, newspaper articles and other bibliographic resources. They are particularly important in academic research and when compiling lists of literature, though they are of course also relevant to anyone searching for information on a particular subject.

Currently it is difficult to gather these data and make them available in a useable form for research or for setting up new digital services. The aim of the working group is for data to be made available for further use as a matter of course. That would be the basis for creating valuable new services for academics and the general public.

Promoting open bibliographic data

What is the working group doing to achieve this?

First of all, we defined the basic principles of what we mean by the term “open bibliographic data”. The group is now concerned on the one hand with clarifying legal questions and choosing appropriate open licences. At the same time, we are working to spread the word, for instance by contacting libraries, publishing houses and other actors and asking them to release their data.

What is more, the working group – like the Open Knowledge Foundation as a whole – is highly practical in its approach. Part of the group, for example, is working on developing simple software to allow bibliographic data to be shared freely on the Web – the beta version was released just a few days ago at http://bibsoup.net. Thus issues we work on also include the various data formats for bibliographic data and how they can be used, edited and interlinked.

An agreement has been concluded that all meta data supplied to the Europeana should be in the public domain. Does that inject momentum into the discussion?

Most certainly. This new agreement on data exchange has for example resulted in all European national libraries signing up to the open data idea. Four national libraries – in the UK, Sweden, Germany and Spain – have already begun to put this into practice by publishing their data, in full or in part, as open data. As a rule, this has been linked to the creation of a linked open data service. All of this is likely to make a not inconsiderable contribution to the further spread of the open data concept in the library world.

Open to new contexts

Linked open data services ensure technical compatibility between data held at different sites. What opportunities do such services present?

Linked open data, or LOD for short, make it possible to contextualize one’s own information – in our case, for example, about books or authors – without any significant technical effort. The data are linked to a different source which publishes information on the same subject.

The advantage is that this allows a service to bring all kinds of information from different sources together, yet the tasks of updating and maintaining the data remain decentralized with the respective services. Library catalogues can then easily display additional information about authors or about the historical period in which a work was produced. Thanks to this sharing of the workload, libraries will be able to continue concentrating on maintaining bibliographic data and information about holdings and their availability.

What do libraries have to do in order to publish their data as linked open data?

If they belong to a library association such as those in the German states of North Rhine-Westphalia, Bavaria or Berlin-Brandenburg, then the publication of linked open data is a service that the association itself will provide. For libraries which are members of the Hesse and South West German associations, this will be done by the university library in Mannheim. In other words, it does not incur any additional work for the libraries themselves – all they have to do is endorse the release of the data.

Innovations for Internet users

Who are the pioneers in this field?

There are many different actors in this area. The first international linked data service – albeit without an open licence – was launched in 2008 on the basis of the Swedish association catalogue LIBRIS. In 2010, the library at CERN, the European Organization for Nuclear Research, was then the first library to publish open data.

In Germany, the central university library (hbz) of the state of North Rhine-Westphalia, where I work, is particularly active. Back in March 2010, it was already setting standards by releasing a large volume of data, and is working constantly to further develop the LOD service lobid.org. Pioneers in the field of linked data also include the German National Library of Economics, the university library in Mannheim and the German National Library.

How quickly can linked open data bring about innovations that the Internet user will notice?

For actors who produce and process data, open data can very quickly offer advantages because relevant data – in whatever format they may be – can be used without any problems. Advantages for end users, however, may take longer to emerge.

I believe that we will start seeing the first noticeable innovations in two to three years’ time. By then it may be possible, for example, that a bibliography in Wikipedia or a hit list in Google will contain a reference to a book that is available in a nearby library.