I’m going to just briefly outline a project we’re working on in the UK right now that we’re pretty excited about. It’s all about developing an open learning analytics framework and basing in on the xAPI.
The project was commissioned earlier in the year, to run for two years, and establish a learning analytics platform for every university in the UK that is ready to adopt Learning Analytics. Lot’s of stakeholders and lots of funding – at least $1.5mm over two years to get this off the ground.
The commisioning organisation is called Jisc, which might be familiar to some of you – they’ve done a lot of work in the EdTech space, especially R&D style projects. Jisc works to make the UK the most digitally advanced education and research nation in the world…
It’s a work in a progress…
But they have a large thread in LA.
The project has three key goals. Improve retention, improve achievement and improve employability.
Jisc believes that with the application of big data techniques, we help institutions reach out to learners who need more help; help students to figure out better paths to success and help produce portable records of experience that really demonstrate what a student can do to potential employers.
This is how they are going to do it.
We’re taking data from a whole range of sources and storing it in the Learning Locker – that’s our Open Source Learning Record Store. We’re actually adding in a range of what we’d call ‘static’ data into the mix; taking into account student information, course information and institution information. This is gathered from Student Information Systems using ETL tools; right now this looks like being Kettle. We’re also converting learning activity providers to use the xAPI. You might have seen we’ve already done this for Moodle. ADL already put out a neat little adapter for EdX. And Blackboard, who play a role in this wider eco-system, are also coming to the party. Jisc is also looking at Library systems and self-declared student data as additional sources.
From the Learning Records Warehouse, data is sent to a couple different systems. The core of these will be the predictive tools provided by Tribal and UNICON / Marist. These are well established tools that will start consuming xAPI data, amongst other data in the LRW, to make predictive analysis about students performance. A student consent service will be used to govern how data is shared and used within the eco-system. This will work with a student app to give learners the same level of insight that administrators already have on their performance; allowing them to set goals and monitor their own performance. Finally, an Alert and Intervention system will be used for push notifications and more. Many of these systems will also use xAPI to feedback further data to the LRW, creating feedback loops through the system.
The group has already been busy. Jisc first published its code of practice for LA, which is available on its website and governs the approach.
We’ve contributed a plugin that takes the core Moodle Activity log, expands on the data stored there, translates it to xAPI and emits it to an LRS of the users choosing. Last time I checked that covered about 20 core actions in Moodle, and more are being added all the time. We did this in partnership with the Charity Learning Consortium, who deserve a big shout out.
We’ve established a multi-tenant Learning Records Warehouse for Jisc, based on Learning Locker. Many of the features we’ve built here will become available in the core product over the coming weeks and months. This includes features like user matching (reconciling users with numerous IDs), and a business rules layer that applies a logic test to incoming data before pushing it out via a Streaming REST API to 3rd party systems.
We’ve mapped the ‘static’ data fields into the LRS, creating the ‘warehouse’ elements that sit outside of the standard API’s. This data varies in sensitivity, so we’ve created different oAuth scopes based on sensitivity that allow only certain access to personal data to other applications.
Finally, we’ve got our first four universities signed up, with a long-list of 30 others looking to push ahead this year.
What we’re creating here will be Europe’s biggest store of educational data. There are going to be opportunities for other providers to get involved, or to learn some of the lessons that we’re experienced in the wider group. If educational LA is on your agenda, we should talk.
This is one of three European funded projects that I know about in Learning Analytics right now. There are probably more. This accounts for some $3mm in funding for the next couple years. Get on a plane, come join us.
Which reminds me. One More Thing…
xAPI Camp, London, will happen on Friday 22nd April, 2016.
Sponsorship and registration details will follow in due course, or come see me or Aaron to know more…
Tea will be had.
Ben Betts Open Learning Analytics and xAPI
Open Learning Analytics and xAPI
Dr. Ben Betts
To enable people in higher education,
further education and skills in the UK to
perform at the forefront of
international practice by exploiting fully
the possibilities of modern digital
empowerment, content and
To make the UK the most
education and research
nation in the world
What do we mean by Learning Analytics?
The application of big data techniques such
as machine based learning and data mining to
help learners and institutions meet their goals:
Code of Practice
Moodle xAPI Plugin
Establish Learning Record Warehouse
User Matching algorithm
& Business Rules layer in LL
Mapped data models for predictive analytics
First institutions signed up
Achievements so far…
Europe’s biggest store of education data will be xAPI
Billions of data points.
App store anyone?
This is one of three European-funded LA initiatives
that uses xAPI.