Data-Driven Library Infrastructure: UKSG 2012 Presentation

Below is a copy of the plenary presentation I gave to the UKSG conference 2012. I have also included a much reduced transcript of the talk to provide some context to the slides.

My presentation was about looking at library services and systems from a data-centric point of view. Specifically, it was about the potential that library data has for the creation of new services and improved systems.

This isn’t a radically new vision – indeed the idea of data-driven is something that seems all pervasive at the moment (data-driven journalism etc). Rather it is a way to refocus, or possibly to re-align our thinking so what may appear problems at the present are viewed as new opportunities.

There is also a video of the presentation available:

I began my presentation with a video. The video was made by University of Lincoln students without formal permission from the university and upoaded to YouTube.

Incredibly it got over 2.1 million hits. Is this the most watched University recruitment film ever!? Even more impressive is that it will be watched by exactly the demographic Lincoln would want to appeal to – young people. Lincoln recognised the potential of the film, and officially branded it and it is now a part of their advertising.

So, I think the film highlights nicely the three main themes of my presentation:

  1. Situating services and infrastructure within the wider ecosystem (this might be institution; community; society etc) – allow innovation to flourish anywhere, and ensure you’re in a position to take advantage of it;
  2. Redistributing effort – focus on the services that have an impact for users, ensure you have the talent to recognise those emergent opportunities and embrace them;
  3. Covering all eventualities: Future proofing –  become agile and more entrepreneurial. The barriers for students creating the video were incredibly low: flip cam and youtube. Barriers to students using library data should be low too
Why is this so important to libraries…? Well, I think there are three compelling reasons why libraries should take a data-centric approach to their systems and services.

1. Ecosystem

Taking a data centric approach enables the library to affect the entire ecosystem that they inhabit.

Focusing on the data forces us to think about the other sources of important data within the institution: the Repository, VLE’s, student records etc.  The wider data ecosystem becomes evident, and the potential of the data underpinning those systems can be realised.

A really good example of this is the Discovery work that’s currently being undertaken by JISC and RLUK and Mimas at the Uni of Manchester. Discovery’s aim is to provide a metadata ecology’ for UK education and research – and it does this by focusing on open and accessible data.

When you start to think like this you realise there is incredibly rich and important metadata describing content outside of HE libraries – museums, galleries, archives, museums. Researchers and students want this content as much as anything in the library – so why wouldn’t you include that?

What was largely hidden or difficult to find becomes visible. The possibilities of those large, cross-sector discovery tools can be realised, as well as those small, un-thought of possibilities for individual researchers to create their own unique discovery tools, searching a corpus of data they curated and specific t their research.

What happens, suddenly, is the data ecosystem starts to mingle with the human ecosystems libraries are inevitably a part of. The free flow of data provides the fertile ground for new ideas and services to grow – Innovation is allowed to flourish everywhere on campus – not just within the confines of the traditional walls of the library.

Libraries and their institutions need to ensure an environment where this flourishing of innovation can happen,  and that there are the right skills and people to recognise those opportunities, and help develop further the ideas and prototypes.

2. Effort

If you think only of shared services as a way to reduce effort then you lose the ability to respond and build on the opportunities that may emerge from your fertile data ecosystem. It’s not of reduction of effort, but of redistribution of effort.

The aim is to reduce the effort on those chore jobs – admin, back office functions – critical, but not what the user sees as having an impact. This refocus allows you to redistribute effort to the core services you provide.

Shared services such as JUSP and Raptor are great examples of how you can stop doing the administration, and use those services to provide you with the data to make changes.  

These shared services also demonstrate the way data begets data: the way data is used produces more data, that enables better understanding of how the data is used and how it can be improved.

Data seems to bestow the need for iterative thinking – to constantly revisit, act, think, revisit.  Providing a virtuous circle of data, action, data.

Another example is Knowledge base+: a shared academic knowledge base for electronic resources – a great example of how you enable libraries to do something once and share it with everyone so they don’t have to repeat it locally. KB+ also recognises that the innovative services built on top of data do not necessarily have to be undertaken by the project itself, but can emerge from the community as well as third parties and commercial suppliers.

One interesting aspect of KB+   is it’s focus on data means that the use cases for it can develop over time. While the focus at the moment is on e-resources and their management, it may enable innovations around inter-library sharing of content, collection management etc.

Indeed, the use cases can quite happily shift from individual institutions:from those that envisage a use-case as an ERM, to those who see it as a backup for their local holdings and helping facilitate easy movement between external systems.

As librarians we’re very aware that the past needs protection from the future; but we need to recognise that the future needs to be protected from the past – I don’t know what will be needed in a few months time; but it probably won’t be the thing I think it will be based on my past experiences.

3. Eventualities

So, this leads me onto my third E! How a data-driven approach ensures that services and systems are Future proofed.

This is about libraries being able to be, at a fundamental level, more entrepreneurial. I want to pick up on an earlier point about the iterative imperatives of data: data – action – data: the process is one similar to how a small start up company might work.

There are some wonderful examples of libraries playing with this kind of innovative approach.  At the University of Huddersfield they have taken the generally unappealing library circulation data and turned it into a game for students:  Lemon Tree.

The library experience is gamified, and in a way that engages students and enhances their experience.

Supporting this kind of thinking are technical infrastructures like the JISC Elevator: A platform for new ideas to be posted, and for members of the community to vote on them and then for JISC to potentially fund.

This is about agility, and moving quickly.

This is again where the importance of redistribution of effort is so essential – the negotiation between shared above campus services and local capabilities is incredibly important. The negotiation between what is shared and kept local defines the institution, and how effective they can be in meeting the rapidly changing needs of their students and researchers.

As libraries begin to understand and curate their own data effectively it begins to demonstrate the libraries potential role in an increasingly data-driven academic environment.

As data management and curation become the next big problem for institutions, libraries can position themselves as the experts.


Leave a Reply