The European Library currently aggregates metadata from 48 national libraries of Europe and over 19 research libraries. The data is searachable via the portal at http://www.theeuropeanlibrary.org/

One of our tasks for 2013 is to create a single linked open data set of this data aggregated. Unsuprisingly, one of the biggest challenges is the related licencing; approaching each libary who is a member to get agreement over the terms and conditions of the linked open data.

From the individual library’s point of view, they need the chance to discuss and ensure that the European Library dataset they are contributing to fits in with their own metadata strategies. From the potential end user’s point of view, they need a data set that is rich as possible and has harmonised licencing.

The success of the work will be in balancing these two viewpoints.

I hope most libraries will be keen to release at least basic metadata as CC0. This follows the trends at libraries such as the British Library and the National Library of Sweden (the National Library of France, meanwhile, has an open licence which equates to CC-BY).

The first issue will be in defining what is meant by basic metadata. Each library might have a different view. Some libraries might be happy to share their entire metadata set; others may only want to release a limited set of elements at CC0

But I suspect the more difficult issue will be the licence terms for data that libraries do not wish to release as basic CC0. Some may choose similar open licences, others may opt for more restrictive terms.

The difficulty then arrives when you then want to put all this metadata together in one unified corpus. In such a case it would need to be released under the more restricted licence – the lowest common denominator. So even if 95% of the libraries gave The European Library to release their full metadata under licence Type A, the complete aggregated dataset would have to be released under the licence Type B stipulated by the other 5%.

Thus the real pressure in this task is trying to convince the libraries to use similar terms for their licencing, creating the richest possible dataset.