October 04, 2004
Stephen Downes - Personalisation of Learning
Stephen talked about Personalisation in Learning. I had asked what can we do now, with current tools?
Download file, 2Mb, MP3, 17:39 m:s, opens in new window.
Well, first of all it can't be done with the tools at hand.
To achieve personalisation in learning, the first thing is to be very clear about what the end point is - Exactly what do I mean by personalisation? Yes, it will be personal, and there will be some points of collaboration and difference. Once that end point is defined then every possible action can be tested against that end-point. This is really basic decision making and goal setting, but served as a useful reminder to define what I mean by 'personalised learning'.
[5:00] Stephen talked about his view of learner centred learning
So, my definition of personalised learning (for the time being):
- personalised learning is about meeting the needs of the learner, with reference to their best way of learning, and their interests. In the world of VET, this can include negotiation about how a learner will learn, what they will learn and responding to their needs and applying to and working in their context. This links into the discussion that followed on Monday 4th Oct about the right of employers to direct learning.
[7:10] How do you present the range of learning opportunties to a person, and create a mechanism where they can select from these opportunites, and then create an environment where they can take advantage of those opportunities?
The short answer is metadata. Or the long answer is a 'Resource Profile' which is between the resource, the learner (eg FOAF style metadata) and the context in which the resource is used. This is my take on the triad concept.
The challenge is, as always, gathering or creating the description. With a description of each of these points, then learners, contexts and resources can be matched up. The metadata should be retrievable by any part of the network. This idea is explored in more detail in Stephen's paper "Resource Profile" here.
The theory is that if you have any two of the triad (generally the learner and the context), then you can make a reasonable selection. Ideally the metadata will be created as the resource is created, is added to when the learner uses the resource including context of use, finally 3rd party metadata would be added in form of classification or evaluative metadata.
Looking to the future this idea would be extended from just-in-time matching, to [14:08] the networked environment suggesting, and adjusting suggestions about resources that might be useful. These suggestions would change over time as the user and/or their context changes.
[17:15] In building a network, the simpler the individual components the better it will work.
Posted by Kirsty at October 4, 2004 07:48 PM