By Andy McGregor November 9, 2011
discovery, libraries, linkeddata

Over the last couple of weeks 3 very interesting reports have drifted through my news feeds on libraries and linked data:

• The library of congress has announced plans for pursuing a replacement for MARC and these plans “will be focused on the Web environment, Linked Data principles and mechanisms, and the Resource Description Framework (RDF) as a basic data model”.
• The W3C library linked data incubator group released their report. This report recommends that librarians experiment more with linked data by releasing data, building on top of linked data sets, engaging with standards bodies and bring their preservation skills to bear on datasets and vocabularies.
• A CLIR report has been published on a linked data workshop and survey run by Stanford. The purpose of the workshop was to discuss the “the prospects for a large scale, multi-national, multi-institutional prototype of a Linked Data environment for discovery of and navigation among the rapidly, chaotically expanding array of academic information resources.” The report itself is useful for everyone as it contains sections on the value of a linked data approach for library content and talks about potential killer apps linked data could support.

These seem significant to me and I am inclined to believe that they represent a growing interest in linked data in libraries. Naturally I have some observational bias in this area since JISC has been funding a fair bit of work investigating the potential for linked library data.

• The Discovery programme funded 8 projects that made metadata openly available, most of these took a linked data approach. Summaries of lessons from these projects will be available very soon from the Discovery website
• Andy Powell and Pete Johnston produced a discussion document on a possible metadata approach to support discovery of library, museum and archive content based on linked data. This attracted detailed and passionate discussion from metadata experts.
• The OpenBib project investigated the issues and possibilities offered by open linked data for bibliographic metadata. We have recently funded this team to build on their initial work to show how this approach to bibliographic data can benefit researchers.
• The ArchivesHub are engaged in a project called linking lives which will use linked data to enable researchers to explore the relationships between people and things that are contained in the archives metadata that the archiveshub aggregates. This builds on the earlier Locah project .
• Suncat are making their journal bibliographic information available as linked data

There is lots of interesting linked data work happening in the wider world of cultural heritage:

• Europeana is taking a linked data approach
• The British Museum is up to some very interesting stuff as part of the Mellon funded ResearchSpace project
• The British Library is engaged in some exciting experiments with a linked data version of the British National Bibliography
• The BBC Digital Public Space project is making use of RDF data to produce a very exciting aggregation of content with many possibilities

This is just a flavour of some of the developments that I am aware of, there are many more, and I don’t doubt that I’ve missed some of the most interesting ones.

So why are so many organisations putting resources into engaging with linked data? Well the advantages of linked data at a very simple level are:

• It enables us to make links between different items in different collections to enable the development of new interfaces that support new ways of exploring collections.
• It can make aggregation and exploration of very different types of data and resources easier
• It works very well on the web enabling clever people to reuse the data to create new tools for engaging with the resources.
• It breaks down the concept of a record of a resource to allow us to make better use of the fields in the record such as people’s names, place names, dates etc.
• It can potentially lead to reduction of duplication of effort if key datasets are shared, this could mean that you just need to link to a trusted dataset rather than devoting effort to creating that data yourself.

However, it is far from certain whether linked data will transform the way libraries work or simply become a tool that is used for some datasets. Many people that I trust still have reservations about linked data as the skills required to model and create linked data are not commonly held by people in most libraries and it is not clear yet that there is an obvious return on the investment required to create and exploit linked data.

My personal opinion is that judging by the amount of effort and work that influential organisations are putting into linked data projects then it is not something that is going away soon. It seems likely that linked data will develop into a useful tool for at least some of the metadata or sets of metadata that librarians use. Senior librarians or those interested in personal development will probably need to think about the skills required to engage with this emerging technology.

As part of the JISC Discovery project we will be dedicating effort to making sure that librarians can learn from the projects we fund to investigate linked data. We hope that this will be a useful learning tool for those with an interest in developing their linked data knowledge or skills. This should include high level messages on value of the approach and detailed lessons on the technical and licensing issues involved. All of our resources will be made available on the Discovery website. We are also planning to provide training on some key topics so keep your eyes peeled for developments.

If any UK libraries are interested in experimenting in this space or in following the innovations of others, they may want to look at our current funding call which makes money available for UK HE libraries, museums and archives to make metadata openly available. There may be just enough time to put a bid together before the deadline of the 21st of November.

Finally, if you are interested in linked data it is worth watching this blog as my colleague David Flanders is planning some further posts to talk about the possibilities linked data offers for higher education.