A Personalized AI Coach to Assist in Self-Directed Learning

Stephen Buckley, John Kos, Rahul Dass, Cathy Teng, Kenneth Eaton, Sareen Zhang, Ashok Goel
Proceedings of the Innovation and Responsibility in AI-Supported Education Workshop, PMLR 273:221-229, 2025.

Abstract

Personalized learning is a powerful tool in online education, yet its application in inquiry-based modeling environments remains underexplored. Previous work has shown that learners that engage in a cycle of construction, parameterization, and simulation, which we refer to as the exploration cycle, create models with higher complexity and variety. In order to further study these findings we present an “exploration coach” that provides personalized feedback within the Virtual Experimental Research Assistant (VERA)—an interactive learning environment for conceptual modeling of complex systems that evaluates models through agent-based simulations. Our architecture, which classifies learners into groups using clustering techniques, allows us to determine what type of feedback would be useful to a learner at any point in their modeling journey. The coach then uses procedural scaffolding to guide learners through the exploration cycle. Lastly we illustrate how these categorizations and the exploration cycle map onto the cycle of self-directed learning.

Cite this Paper

BibTeX
@InProceedings{pmlr-v273-buckley25a, title = {A Personalized AI Coach to Assist in Self-Directed Learning}, author = {Buckley, Stephen and Kos, John and Dass, Rahul and Teng, Cathy and Eaton, Kenneth and Zhang, Sareen and Goel, Ashok}, booktitle = {Proceedings of the Innovation and Responsibility in AI-Supported Education Workshop}, pages = {221--229}, year = {2025}, editor = {Wang, Zichao and Woodhead, Simon and Ananda, Muktha and Mallick, Debshila Basu and Sharpnack, James and Burstein, Jill}, volume = {273}, series = {Proceedings of Machine Learning Research}, month = {03 Mar}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v273/main/assets/buckley25a/buckley25a.pdf}, url = {https://proceedings.mlr.press/v273/buckley25a.html}, abstract = {Personalized learning is a powerful tool in online education, yet its application in inquiry-based modeling environments remains underexplored. Previous work has shown that learners that engage in a cycle of construction, parameterization, and simulation, which we refer to as the exploration cycle, create models with higher complexity and variety. In order to further study these findings we present an “exploration coach” that provides personalized feedback within the Virtual Experimental Research Assistant (VERA)—an interactive learning environment for conceptual modeling of complex systems that evaluates models through agent-based simulations. Our architecture, which classifies learners into groups using clustering techniques, allows us to determine what type of feedback would be useful to a learner at any point in their modeling journey. The coach then uses procedural scaffolding to guide learners through the exploration cycle. Lastly we illustrate how these categorizations and the exploration cycle map onto the cycle of self-directed learning.} }
Endnote
%0 Conference Paper %T A Personalized AI Coach to Assist in Self-Directed Learning %A Stephen Buckley %A John Kos %A Rahul Dass %A Cathy Teng %A Kenneth Eaton %A Sareen Zhang %A Ashok Goel %B Proceedings of the Innovation and Responsibility in AI-Supported Education Workshop %C Proceedings of Machine Learning Research %D 2025 %E Zichao Wang %E Simon Woodhead %E Muktha Ananda %E Debshila Basu Mallick %E James Sharpnack %E Jill Burstein %F pmlr-v273-buckley25a %I PMLR %P 221--229 %U https://proceedings.mlr.press/v273/buckley25a.html %V 273 %X Personalized learning is a powerful tool in online education, yet its application in inquiry-based modeling environments remains underexplored. Previous work has shown that learners that engage in a cycle of construction, parameterization, and simulation, which we refer to as the exploration cycle, create models with higher complexity and variety. In order to further study these findings we present an “exploration coach” that provides personalized feedback within the Virtual Experimental Research Assistant (VERA)—an interactive learning environment for conceptual modeling of complex systems that evaluates models through agent-based simulations. Our architecture, which classifies learners into groups using clustering techniques, allows us to determine what type of feedback would be useful to a learner at any point in their modeling journey. The coach then uses procedural scaffolding to guide learners through the exploration cycle. Lastly we illustrate how these categorizations and the exploration cycle map onto the cycle of self-directed learning.
APA
Buckley, S., Kos, J., Dass, R., Teng, C., Eaton, K., Zhang, S. & Goel, A.. (2025). A Personalized AI Coach to Assist in Self-Directed Learning. Proceedings of the Innovation and Responsibility in AI-Supported Education Workshop, in Proceedings of Machine Learning Research 273:221-229 Available from https://proceedings.mlr.press/v273/buckley25a.html.

Related Material