Activity data is big business. We see it in the recommendations we get every time we look at something on Amazon, we see its importance every time we get asked if we have a club/nectar/loyalty card when we buy something and we see it in the fascinating story of the Netflix million dollar prize to improve film recommendations for their users. Higher education institutions have all sorts of data stores about the activity of their employees and students. Are there ways that this data can be used to improve the research and learning experience?
This was the question that JISC set out to answer by funding 9 experimental projects to analyse and exploit activity data to provide new services to researchers and students or to improve existing services. These projects covered recommendation services for library and repository content, access grid usage patterns, analysing data for student retention, virtual learning environment usage data, the link between student attainment and library usage and the possibility of taking a user centered approach to activity data.
The simple answer is yes, there are lots of ways that working with activity data could be useful. But there are many technical, legal, skills and policy issues that need to be addressed in order to do so. To enable others to learn from the experience of these projects and to copy the technical and legal solutions they developed we have produced a site that summarises all that was learned in the programme. The site is live now. It provides a high level overview of what activity data is, why it is useful and how it can be exploited and also contains detailed recipes for anyone who wants to start the process of exploiting activity data at their institution. You can also read more about the projects that made up the programme.
The site was produced by Sero Consulting working with Tom Franklin and Mark van Harmelen.
One interesting question about activity data is should it be made openly available? There are plenty of challenging issues here to do with anonymisation, compliance with data protection and ensuring that the users are appropriately informed and it is what they want. But there are also benefits to open data. A good illustration of what can happen with open activity data is the Book Galaxy app below that uses the library circulation data released as part of the JISC Mosaic project to provide an innovative interface for exploring the relationship between books. The activity data guide includes resources that address the question of open data and links to open data released by some of the projects.
Here’s the Book Galaxy – move your mouse around the galaxy to see the titles. Blue dots are books, yellow dots are courses. If nothing is appearing, you might not have Java installed. Download Java
It seems to me that there is likely to be appetite for further innovation with activity data as it offers the potential for more efficient institutional services and new functionality that can enrich the research and learning experience. Both of which are important drivers in the current climate. We will be funding further work on 4 of the 9 projects to explore whether they can develop further answers or produce useful services. There is also a programme of projects on Business Intelligence managed by my colleague Myles Danson, these projects are building solutions for storing and analysing data about the business critical operations within universities. You can read an overview of the projects on the JISC website and there is a useful JISC Infonet Infokit on the topic of business intelligence. Within JISC we are also starting to think about the bigger picture of business intelligence for universities and what we can do to help universities exploit emerging opportunities so look out for future work in this area.
Find out more about this work and that of the digital infrastructure team on their blog.
Hi David,I think we used the phrase Blended Learning, to make sure that staff think about the rnnlaieg experience and not only the teaching experience.I’d be worried by a phrase like Blended Teaching because it might make it look as if the purpose of the activity is teaching. I subscribe to the view that teaching is not a good in itself but an enabler to rnnlaieg. Thus producing a good blend of processes and technologies in our teaching is only worth doing if it enhances the student experience and helps them improve their rnnlaieg. Thus blending our teaching is, in this view, Blended Learning.
Another important way that data is oparhned within organizations is by being locked up within documents; the USA government’s practice of publishing data tables as PDF is an example!The linked data model also includes the co-publication of documents as data, e.g. through RDFa.