Focus on research: Owen Conlan, Adapt
29 November 2016 | 0
Owen Conlan is an associate professor at the Science Foundation Ireland-backed Adapt centre for digital content research based at Trinity College Dublin. Here he talks to TechCentral.ie about the role of personalisation in education, how linked content will change the way we see the world and why risk should hold no fear for students.
People are used to the concept of personalisation in how websites serve up content or make recommendations for things to buy. What does personalisation mean in a classroom setting?
The personalisation offered through online shopping sites and mixed through social media feeds is very limited. In Adapt when we look at personalisation we’re crafting bespoke experiences that engage and stimulate people. This involves weaving together content and services that make sense for individual users in terms of their goals, prior experience and current context.
However, it’s important to remember that the user is not a passive recipient. They should be able to engage with what’s being offered and change things up to meet their needs. This is what we learned from developing personalisation for use in schools. Every student comes with a different perspective on the material being offered.
The personalisation environment Cultura allows students to move seamlessly between exploring material on the Irish Rebellion of 1641 and reflect on what they’d learned. It offered personalised recommendations – similar to how an online store does – but, importantly, it explained why recommendations were being given.
When a student explored concepts they were unfamiliar with, Cultura would offer a personalised expert-authored guide that could take them through the new material in steps, grounding it with concrete examples.
The reflection component was the bit that had the biggest impact on the students. They could see what the system believed they were interested in. Moreover, they could change those interests, and the recommendations and guides would adjust to meet their needs.
Looking at the Cultura project as an example of how to engage with specific events, what challenges did this present?
Even when proposing Cultura I knew we had a fantastic opportunity to engage with a variety of different groups, from members of the public to professional historians.
One of the keys to doing this was to not think of these as very different groups… who says a professional historian knows more about the 1641 Rebellion than someone who grew up near the site of an infamous massacre from that year (or during the Cromwellian Invasion)?
The challenge for us was that the personalisation had to handle a spectrum of users’ prior knowledge, experience and needs. We achieved this by offering a range of exploration tools, scrutable recommendations, tailored guidance and reflective visualisations.
Another key challenge we addressed was in attempting to foster serendipity. This sounds like a contradiction and to a certain extent it is, but it was important that users didn’t get caught in a filter bubble, ie only seeing things related to a narrow set of queries. This is particularly important when trying to get an overview of historical events, where objectivity and critical analysis is important.
In your work you’re looking at the idea of moving from metadata to linked data. How does this process work and what kind of insights is it yielding?
For knowledge-based systems, such as personalisation systems, to appear informed it is important that they can draw connections between related concepts and between content that represents those concepts. This is the key opportunity presented by linked open data, where we no longer have standalone metadata describing individual content resources, but have a web of connected concepts.
For many years personalisation systems were closed-corpus systems only capable of delivering experiences across a (relatively) tiny corpus of content and documents. Now our systems can branch out into the Open Web drawing on structured information from anywhere. The diversity of potential personalised experiences is now almost endless.
Having managed large scale projects with significant EU backing, do you think there is scope for students to learn about project management as part of an undergraduate degree?
Absolutely. In fact we’re seeing that in a large number of STEM-related degree programmes, where undergraduates are exposed not only to the theory and concepts of project management, but are often performing group projects where they can employ develop PM skills.
Managing research projects is mainly about managing risk. Some of that risk can be mitigated, but it there always needs to be a ‘Plan B’. It would be great if we could find ways for undergraduate students to experience how to manage risk.
Where do you see personalisation research branching out to next?
Personalisation has the potential to help unlock our potential! Imagine brainstorming with a technology that offered ‘what ifs’ based on your interests, was as agile as you are and could help you sort the good ideas from the weak ones. This is the potential of personalisation technologies.
We currently experience personalisation as flat and intrusive advertising (often trying to sell us something we’ve already bought or aren’t interested in). With the work we’ve done in Adapt we’ve developed deep and meaningful personalised experiences where the user is in control.
The next step is to develop technologies that help people to be more creative, rapidly explore new ideas and to unlock more of our potential.