We’re asked questions like this regularly, by customers, developers, learning experts, evangelists and blog readers. At Learning Pool, we’ve had a secret plan over the last few years to give our LMS more intelligence, and we thought it best to sit down and put these questions to the man behind the plan, Learning Pool’s Technical Architect, Ryan Adams.
Ryan, is it possible to efficiently create a course that lets people learn what they need to learn?
Adapt gives us granular content with a clear structure that lets us re-use and report on individual segments of the module. The raw Adapt content is in an open format, which allows us to write software that understands the information held within the course. We knew that cloud-based authoring would let us create relationships between modules – something that we couldn’t do with the desktop authoring tool. With Adapt Builder , we can now keep track of the ancestry of modules, so we know when content is shared and remixed across many LMS.
We knew that xAPI could give us detailed analytics about the behaviour of our learners. We knew that it opened up the opportunity of recording learning that happened outside the LMS. We saw the potential in its offline capabilities and the possibility of capturing more specific records. We knew that learner records were locked up inside the LMS and that we needed a Learning Record Store to allow us to aggregate and analyse records from all our learning systems. We’re rolling this out in the next few months.
So if we have a record store with several hundred thousand user records in it, what can we do with all that information?
Magic, that’s what.
If we have a collection of courses that are structured, machine readable and have a shared inheritance; and information about learning that takes place outside the LMS, how about an LMS that suggests the next thing you need to look at based on what other people in your situation did? A tool that strips out the bits of a course that you don’t need because you know them already, letting you get on with work faster. A learning product that knows how you’re performing in your job and suggests things you might need reminding of. With this we know what it is they need to learn, what they already know and making sure that the course we create is as good as it can be.
Imagine an authoring tool that highlights sections that learners will struggle with. One that suggests extra content to improve your course. That warns when your course is too long and learners are failing to complete it. What about a system that sees when user behaviour isn’t changing as intended and suggests some Encore activities?
Wow, sounds great! Are there pitfalls and problems that we know we need to avoid?
We need to take care with privacy and confidentiality. We need to keep your organisation’s content and staff records secure while at the same time using them to train our recommendation engine. We have a robust permissions model in Adapt Builder that will apply to our LRS. Did you know we’re ISO 27001 certified?
With any recommendation tool it’s possible to generate inappropriate suggestions. How many of us have had the pair of shoes we just bought follow us round in an ad on every website we visit? It’s also easy to get stuck in a feedback loop and get the same things recommended all the time. We may need to make room for some serendipity and randomness in our suggestions. We also need allow for variations in the content or style of the content we recommend; it’s all too easy to produce personalised content with no personality.
So when are we going to see more predictive learning hosted on a more intelligent LMS?
Any predictive analysis tool is only as good as the data you feed it and the algorithms you build into it. With almost a million learners and our huge catalogue of Adapt Modules we have an excellent base to begin with.
Over the next few months you’ll begin to see these systems roll out; a new catalogue product that’s integrated with Adapt Builder and your LMS; xAPI support in Adapt modules out of the box; and an LRS integrated with your LMS.