Behavioral data analytics in intelligent tutoring systems

This project aims at developing automated immediate and personalised feedback for a smart learning environment used for teaching conceptual modelling, an ill-structured complex learning task. Addressing this challenge will make teaching and training more efficient, provide an education better adapted to each learner, and advance the theoretical underpinnings in digital learning technologies. The project's core innovation is three-fold: 1) we will build novel predictive process analytics techniques that, taking into account the behavioural learner data, will automatically detect whether and when feedback is needed by an individual learner; 2) we will develop a generic theoretical framework that will give insights and recommendations about feedback automation and will guide the design of a generic architecture for intelligent and automated provisioning of feedback; 3) a state-of-the-art in-house learning environment will implement this architecture and automatically deliver the needed personal and immediate feedback by exploiting the predictions from the process analytics and the theoretical framework guidance. Experiences from this case study on conceptual modelling will be generalizable to other digital learning environments. The specific goal of this PhD project is to investigate the development of predictive process analytics techniques for building models that are capable to detect feedback interventions in real-time.

Publication year:2021

Accessibility:Open