Erasmus+ project led by Oulu University used a chatbot to personalise higher education students’ learning experience

The idea of individualising studying with the help of chatbots originally emerged in Oulu already around 2019 as the university carried out its first experiments relating to this theme together with the IBM.  

“The idea of using chatbots stemmed from a need we had discovered when teaching project management to industrial management student. We wanted to find out how we could support students’ learning more actively in the online world, which we were already introducing before the COVID-19 pandemic,” explains Kirsi Aaltonen, Professor at the University of Oulu. 

Chatbots had already been used for student counselling, but we now started testing them in the actual instruction and efforts to enhance its impact. 

“We wanted to develop this idea in an international environment together with professors and researchers of project management. We had already worked on teaching development together with some of the participating universities.” 

At the start of the project, chatbots as a teaching aid were an uncharted territory, and they were hardly used at all for teaching actual content. While such actors as the IBM had already carried out some experiments, they were more relevant to learning processes. 

“From the beginning, we set out to consider both the learning process and ways in which technology can support learning,” says Hannele Lampela, Senior Research Fellow at the University of Oulu. “We expanded the project’s cooperation network by involving research in education science, and especially research that combines learning and technology, conducted at our university. Rather than only developing a technological solution, we wished specifically to improve the student’s learning experience.” 

This way, the learning process and technological development progressed hand in hand in the project. 

As it happened, the project coincided with a period of rapid development in the AI field. Large language models, generative artificial intelligence, and ultimately ChatGPT in November 2022 were part of the picture. All this made ChatLearn an even more topical and interesting project.  

Teaching chatbots can also facilitate teachers’ work

One of the aims of the project was to make teachers’ work easier. Large courses and mass lectures are attended by many students with different backgrounds and capabilities, and providing individual support for them is challenging, if not impossible, for the teacher. 

One such subject taught to large groups of students is project management. Skills in this area are currently taught in most fields of science, including industrial management and engineering. Students may find studying certain project management techniques difficult, which is why these specific teaching contents were selected for the project. In addition, building a technical learning environment for this type of content is more challenging than for mathematics or other natural sciences, for example. In the latter fields they have been used for several years, with Ville mathematics assignments for basic education as an example. 

Project management learning content which students had found difficult and in which more individual learning support was consequently needed was incorporated in the chatbot devised by the ChatLearn project. In practice, the chatbot platform provided the student with information on different project management topics as well as assignments and more open challenges related to project management to complete. If the student's answers were inadequate, the chatbot provided them with suitable content-related support and additional materials that helped them complete the assignment. 

“Feedback from students and teachers was almost exclusively positive,” says Lampela. “They also provided valuable knowledge about how the chatbot could be developed further. 

The project outcomes and experiences of using the teaching chatbot have been disseminated both in scientific articles and at various events and webinars. 

Demand for developing teaching chatbots exists

Since those days, the use of AI as a learning tool has become somewhat more common in the academic world. Especially during the past year, international universities have introduced new kinds of platforms and applications that enable universities and teachers to develop AI solutions suitable for their specific teaching content. 

“From a pedagogical viewpoint, however, many of these applications are not yet very progressive,” notes Aaltonen. “Most of the published research has also focused on what ChatGPT can offer for instruction or what kind of risks its use involves. In itself, however, it is just one application, and there is a need for research and applications that enable better personalisation of teaching.”

Lampela also points out the need to address accessibility in AI solutions. This poses its own challenges, which this project did not have time to address particularly thoroughly.

There is a demand for research in teaching chatbots and innovative learning applications that make better use of AI in the higher education world. After all, Digivision work also aims to promote digitalisation in higher education. 

“It was a positive surprise for me to witness the high impact that such pilot projects and experiments can have on advancing thinking and coming up with ideas for developing digital solutions. Personally, I would say that such teaching chatbots may play a major role in how we can optimally personalise teaching and support learners in the future. We are only taking our first steps on this development path,” Aaltonen notes.

There are already plenty of possibilities – from the perspective of both content and pedagogy. In terms of advancing this development, more positive attitudes towards experiments and digitalisation in the higher education sector will be crucial. 
The chatbot developed in the ChatLearn project is no longer directly used in its original form, but the materials and contents of the project are openly available on the Erasmus+ Project Results Platform. Anyone can build their own teaching chatbot application on their basis. 

Text: Aino Kivelä