An interview: Gábor Kismihók, head of the TIB junior research group “Learning and Skill Analytics”
How will we learn in the future? Or: What will learning and education look like in a few years? These are precisely the questions you want to answer with your research.
These days we are experiencing a radical change in how we learn and teach. With the help of intelligent technologies we can now personalize the learning experience of each and every learner, which was not possible in the past. We can consider a number of contextual factors, like individual goals, geographical location, professional and educational achievements and provide learning assistance to every learner accordingly. This is a very exciting area at the moment both from scientific and application point of views. With my research group we want to contribute with methods and tools to this notion and help the further democratization of education.
This work needs a number of complementary expertise and teamwork. Therefore, in my group we have experts from the areas of educational psychology, computer science, data science, and statistics. My own background is in business information management, and I’m specialized on information systems in the area of learning and education, which has been a very dynamically developing segment of the world of information systems.
Since I started to work on this topic in the mid 2000’s (during my doctoral research), this area has seen a lot of novel and exciting developments starting from virtual classrooms and other digital learning environments, to the current trends of Artificial Intelligence based applications. I’m happy to see that educational technology is now in a daily use by learners and educational institutions, and, especially in these COVID times, often proved to be useful and effective.
A really exciting topic. What are the opportunities and challenges?
Yes, the digitization and subsequently the personalization of learning is a very exciting topic indeed. I believe that our societies are getting more and more involved in the exploration of digital world for learning. Addressing individual learning objectives, focusing on personal skills and career development, providing quality education to those who can’t afford expensive schools and universities, but still have access to basic technical infrastructure (e.g. a mobile phone), are benefits we should always keep in our sight.
However we still have a large number of challenges to solve, here I want to highlight two of them. The first one is about capturing the learner’s context. Identifying, storing and analyzing contextual factors in learning in an open and transparent way are not obvious exercises yet. Oftentimes they require the management of sensitive personal data about learners, so we need to be very mindful and careful when collecting, storing, analyzing and visualizing these data.
The other one is to keep learning human. Learning has to be a (mostly) enjoyable and personal journey, where we oftentimes have to interact with others and with intelligent technology. I think a major challenge ahead of us is to find out how to create such highly digitized, AI driven learning environments, where we can safeguard the human aspects, the human touch of learning in . Actually, one of our projects is addressing precisely this issue: OSCAR is trying to match AI based learning recommendations with human coaching and mentoring in a single digital learning environment.
Which research topics are you currently working on?
Our flagship project is eDoer, in which we want to show how AI can help individual learners to improve their work related skills. This is done by developing open and transparent algorithms, with the subsequent utilization of the vast amount of globally available Open Educational Resources (OERs) on the internet. These OERs are offered to learners according to their personal learning preferences, so they can build their own learning pathways for their career goals.
The process is simple: First, with our recommender, learners can set their career goals, and retrieve the necessary skills for their learning objectives. Afterwards, they generate a list of learning topics they should master for skills they want to focus on. Once this is done, we recommend OERs for all the selected topics, what are combined with assessments to assist and monitor the learning progress.
This project is in the sweet-spot between academic and applied research. In order to make this concept work, we need to develop new algorithms and concepts on how such an Artificial Intelligence powered recommender should work and interact with learners. Our first prototype is already up and running for jobs related to data science. Please try it – under http://edoer.eu/ – and send us your feedback! Any help on our approach is more than welcome!
At the beginning of 2021, two more projects start, in which you and TIB are involved – ADAPT and BIPER. Congratulations! What are the two projects about?
Thanks! ADAPT will be an excellent use case of the eDoer platform. We will fine tune our learning content recommender system for nursing and care taking jobs, so those people who work in these areas will be able to train themselves efficiently according to their own precise training needs. In this project TIB will be responsible for delivering and managing the intelligent training platform. ADAPT is financed by the BMBF, it will last for 3 years, and we will be able to recruit 2 new researchers for this adventure.
BIPER is a much smaller and more conceptual project. It is funded by the Erasmus Plus programme of the European Commission, and coordinated by the Corvinus University of Budapest – which happens to be my alma mater too. In the upcoming 1.5 years we will look at how can we further personalize and digitize current educational curricula in the area of Business Information Systems.
This interview first was published in the TIB Annual Report 2020.