Teachers and schools are under increasing pressure to surf the wave of digitalisation that is currently dominating the rest of our society. School management systems such as Smartschool are no longer part of the daily school context. However, in the daily classroom practice itself, there are still many opportunities for the deployment of technology and digital tools.
So do we just switch to online education?
As a society, we expect our teachers to gradually make the transition to a form of blended learning, in which traditional teaching is supported by digital tools. Easier said than done… The corona crisis has indeed accelerated the whole digital learning issue, forcing teachers to embrace technology and switch almost entirely to online teaching. Really punishing how all those teachers changed their teaching style overnight. Hats off!
But how do we make this sustainable in the long run? Teachers are currently busy 24 hours a day, so to speak, preparing their lessons, recording videos, testing educational applications and improving their pupils’ online and offline assignments. This is unsustainable in the long run. It is therefore important that we use the current situation as a learning opportunity: this corona crisis has shown that the possibilities of educational technology are endless. Moreover, we can now see where extra expertise is needed so that educational technology can be deployed in a sustainable manner in future, without overloading our teachers.
A guide to digital tools?
For many teachers – especially those who did not grow up with technology – the exploration of educational technology is a titanic task. Just look at the proliferation of distance learning tool overview websites. The range of educational tools on offer is now overwhelming. As a teacher, how do you decide which educational tools are good and useful for your pupils? The visitor numbers of tools such as www.onderwijsgaatviraal.be (in Dutch) do not lie: the need for support is great! Within the i-Learn project, coaching will therefore also form an important pillar.
Recommendation systems as an aid to choice stress
Choice stress is of course not limited to the world of educational technology. To cope with a sometimes overwhelming supply, tools are being set up in other contexts to help the user make a choice. For example, when choosing a film or series, or when buying a product online. Computer scientists have developed recommendation systems for this purpose: thanks to these systems, you receive free suggestions about which jumper might look good on you and which TV series definitely deserves your attention.
Choices are usually not made on our own. We ask our friends or family for advice. Referral systems also work this way, giving you suggestions based on the choices of users who are similar to you. For example, streaming service Netflix gives you a list of films and series that you might like. This list is compiled based on two parameters: (1) your own viewing behaviour, and (2) the viewing behaviour of similar viewer profiles. In itself, setting up a recommendation system for films and series is not too complex. Films can be easily divided into genres and Netflix assigns a genre and thematic tags to each film or series. Based on this, it is relatively easy to find similarities between different series and films and thus make an educated guess as to what you, as a Netflix subscriber, might like.
In education, however, the issue is somewhat more complex as finding similar “users” (i.e. learners) and educational tools is much more difficult due to the wide variety of tools and learning styles. A recommendation system for educational technology would therefore be a dream for teachers. (Provided that we take some ethical objections into account, and that we can ensure an optimum interaction between the role of the teacher and the support of the recommendation system).
The i-Learn project and the development of our online portal for digital personalised learning should therefore also pay attention to this specific issue. Researcher Sohum Bhatt focuses in his doctoral research on the possibility of developing a recommendation system for learning content. His research results will be taken into consideration in the further exploration of possible functionalities of the i-Learn portal.
How learners learn
Learners each have their own way of learning that suits them best. Some like to read and tell stories, others are more comfortable building all kinds of structures in LEGO, and still others benefit from social interaction when working on a project together with classmates. All these differences between pupils can be described using learning styles.
Let us consider, for example, a pupil who always reads the introduction to each chapter of his physics textbook. This learning style implies that the learner in question likes to read learning content in order to understand and acquire in a rather passive way. Another learner might turn to an app to do a physics experiment. This indicates that the latter student is more open to active experimentation and learns a lot from visual stimulation. These observations and also additional information (such as processing time) can be included in a categorisation system.
A Netflix for learning content?
Let’s take some inspiration from Netflix. If you like thrillers, Netflix will fill your home screen with suggestions for thrillers and films that are a bit out of your comfort zone, but are likely to pique your interest. Netflix does this by mapping your viewing habits to create a viewer profile for you.
In theory, we could do exactly the same with learning. By observing a learner’s behaviour online, we can compile a learning profile. Based on that, we can then easily offer learning content that suits him or her and explore the limits of his or her ability. With the help of an EdTech recommendation system, for example, we can drastically reduce teachers’ planning burden. The system, together with the teacher, gets to know the students better, while the final decisions are still made by the teacher. This could be a useful tool for the teacher, who of course will still play the main role in the pupil’s learning process. However, the personalisation of each learner’s learning pathway could be partly automated, freeing up time for the teacher to provide customised support.
A recommendation system for educational applications could therefore make life much easier for teachers. Thanks to the learning behaviour of students online and the educational applications they use, we can gradually find out which type of learner the tool in question is best suited to. In this way, a classification system of educational tools will emerge in the background that is itself constantly learning and becoming more comprehensive. This will mean that in future a particular application will more quickly be presented within the learning path of a similar learner. However, it is still up to the teacher to choose from these proposals which educational tools are best suited to a particular learner.
By assigning not only a subject area to educational applications and learning activities by means of a tag, but also a learning style, a recommendation system can be developed within a portal environment. In this way, a teacher will see a list of recommended tools and activities for each pupil, adapted to the individual learning style of that pupil. A Netflix for digital personalised learning could be within reach.
Sohum’s research will undoubtedly lead to interesting results in the coming months and years. We look forward to his findings. Thank you, Sohum, for giving us a hint!