Are you a researcher with a passion for advanced machine learning techniques, data science, and Bayesian modelling and inference? Do you want to work in an interdisciplinary team, conducting research in a pleasant and open environment? As a postdoc, you will study AI technologies in relation to self-regulated learning. In this way, you generate insights in the type of hybrid support that is best suited for learning.

We are looking for a postdoctoral researcher with a background in Artificial Intelligence or a related field to work on the Hybrid Human-AI Regulation (HHAIR) project for 4 years (0.6 FTE) or 3 years (0.8 FTE) (depending on your capabilities, the position may be scaled up to 1 FTE with teaching duties).

Young learners (10-14 years) in today's society often use Adaptive Learning Technologies (ALTs) such as Gynzy, to optimise their learning experience of, for instance, mathematics and languages. However, existing ALTs take the control of the learning process away from the student. As a result, students have less opportunity to practise their self-regulated learning (SRL) skills. SRL involves students' control and monitoring of their own learning process, with metacognitive activities such as orientation, planning and evaluation.

In the Hybrid Human-AI Regulation (HHAIR) project, we gradually transfer ownership of the learning process from AI-regulated back to self-regulated learning, which we call hybrid learning. This way, young learners can benefit from an AI-assisted learning programme, while also developing their SRL skills. Depending on their SRL skills, the ALT will adapt to their developing needs. As such, HHAIR facilitates optimal learning of skills. Based on the ALTs trace data, we will determine the level of SRL a learner shows and the amount of AI guidance needed. The project is the first to explicitly combine SRL with AI-supported learning.

As a postdoctoral researcher, you will be part of the ALL team that investigates the ALTs trace data from daily usage in schools across Europe provided by the AI@EDU infrastructure. This process starts with an exploratory step (WP1). Here, we use Bayesian nonparametric modelling to characterise different types of learners and their prototypical behaviour. Subsequently, the insights from WP1 are incorporated into design studies. We use the results from the Bayesian model to predict which type of hybrid support is best suited for a student's learning (WP2). Together with our team, you will be responsible for the development and implementation of these algorithms.

The HHAIR team consists of Dr Inge Molenaar (Educational Science), Dr Max Hinne (Artificial Intelligence), Prof. Eliane Segers (Educational Science), two PhD candidates and one postdoctoral researcher.

The full project description is available upon request.