With so much buzz around concepts like big data it isn’t surprising that librarians are increasingly interested in how they can use library data to help develop improved library services and systems.
Indeed, librarians find themselves in an enviable position. As a matter of course libraries collect a host of usage and transactional data created by users as they interact with their systems and services. They are awash with this type of data –and are waking up the potential value that can be extracted from what at the moment is largely, unstructured data.
Librarians are often at the forefront within institutions in both recognising the opportunities the data presents, and acting on this data. They are helping inform collection management decisions, designing new forms of interactive, game-like participation and collaborating in strategic institutional initiatives in student learning analytics.
But, rather than simply getting caught up in the promises of big data,I want to employ the concept of ‘small data’ - the localised, contextual and manageable data that can help provide a fertile environment for the development of data analysis.
In this post I want to use ‘small’ data as a way to explore some of the analytics developments I’m seeing in academic libraries. In particular this post will highlight some of the ways libraries are making effective use of their data to create better user experiences and inform the development of new services and systems.
Data: big and small
Big data is commonly thought of as a collection of data sets that are so large and complex they are difficult to process using standard database management and processing tools. Typically, big data is applied to areas such as physics or astronomy where the data collected from experiments and observations is enormous, as is the attendant burden of storage, analysis and visualisation.
In contrast to both the scale and hyperbole associated with big data is the concept of small data. Rufus Pollock, of the Open Knowledge Foundation, describes small data as ‘the amount of data you can conveniently store and process on a single machine, and in particular, a high-end laptop or server’.
Pollock contrasts small data explicitly with big data; it’s the data you can manage locally using your everyday tools. What’s important here is that the data originates from a localised context where it might be enriched and improved, analysed and acted upon.
Realising the potential of small data
Small data is inherently local, distributed and context-rich. Given these qualities some academic libraries have already sought to take advantage of this type of data.
Projects such as the Library Impact Data Project at Huddersfield, Library Cube at Wollongong in Australia and Surfacing the Academic Long tail(SALT) at University of Manchester, supported by Jisc, are trying to understand the potential of local data sets - from usage data to bibliographic metadata – and how it can enhance the student and researcher library experience.
The first two projects aim to understand the impact that libraries have on the attainment of students. The latter attempts to go further, exploring the potential of how catalogue metadata might help inform the learning/research process of humanities and social science in new ways, by surfacing material that would otherwise not have been discovered by researchers.
Small data also has the potential to transform the everyday and mundane into something new – and fun. Take Lemon Tree, which has ‘gamified’ the library experience at the University of Huddersfield, and BookedIn a similar initiative at the University of Manchester, by turning book-borrowing and the use of e-resources into a social game where you can compete against your friends and fellow students.The game is beginning to be used in other libraries, both academic and public.
When this small, localised data is aggregated at a higher level (by region, sector or nation) that the various data sets become potentially more powerful and the opportunity for new insights and types of engagement is really unlocked.
In this area Jisc and the University of Huddersfield have begun developing the Library Analytics and Metrics Project (LAMP) which will develop a shared, national service to enable institutions and libraries to make sense of their disparate and diverse data sets.
This should enable them to spend their precious time and resources acting upon it to improve the student and researcher experience. The project is ingesting a variety of library and institutional data in order to clean and normalise the data and present it back to the library via an intuitive data dashboard.
Can services like LAMP free-up library expertise to focus on getting the small data right, so that it can feed into big data services and provide more meaningful, context sensitive and rich analysis? Can we leave libraries to focus on the areas that they already have skill sets in, and provide an infrastructure and tools that makes exploiting big data as painless as possible?Projects like LAMP are attempting to understand how much of the ‘heavy-lifting’ around data collection, clean-up and analysis can be done on behalf of the library, and how much has to be done locally.
A data-driven future?
The future of libraries may depend, in no small part, on their ability to analyse and act upon the various types of data they both generate and have access to, and to reconnect with students and users, in innovative new ways. This future will depend on new types of roles embedded within the library and new types of services. It will mean using different types of data – from numerical data through to qualitative observations - to inform decision making across the library.
For example, understanding the information seeking behaviours and motivations of students is arguably more complex and fluid than what we might be able to get from quantitative data analysis alone.
It may be that the future of libraries is about engaging more deeply with the user experience and includes approaches like those adopted in some US institutions, where anthropologists are employed and embedded within the library. Donna Lanclos at UNC Charlotte is an example of a library ethnographer embedded within the library and who blogs at Anthropologist in the Stacks. Donna’s role ensures the decisions the library makes about service, systems and space are all anchored by the behaviours, practices and motivations of the students at UNC Charlotte.
Such an approach may seem less feasible within a UK context; less budget, lack of skills, and so on. However, the opportunity for libraries to offer internships to local anthropology students and researchers seems one that could enable libraries to embed this kind of understanding of student behaviours deeply within their institution.
If libraries are going to be able to meet student expectations and enhance the student experience, then they will need to be in a position to use data to develop the kinds of services and systems that students and researchers need and want to use. Small data provides libraries with an opportunity to collect, analyse and act upon data that they have and that can make a meaningful difference to their users. It’s not really about big data. It’s about how the small data can be acted upon, and how it can be brought together with other small data to create more nuanced and insightful analysis.
How can libraries use big (and small) data better? What innovative uses have you come across? Let us know in the comments below
Related knowledge and skills