Naaz Sibia, PhD candidate, Faculty of Arts & Science

Supervised by Michael Liut, Faculty of Arts & Science, Tingting Zhu, University of Toronto Mississauga, and Carolina Nobre, Faculty of Arts & Science

Project Title: Scaling Self-Explanations in Online Learning Environment Systems

Project Summary: The pandemic-induced transition to online learning platforms revealed a trend among students towards superficial learning methods, such as accelerated video playback and haphazard answer attempts, often accompanied by environmental distractions. This project aims to bolster pandemic readiness and educational resilience by fostering deep learning through large language model (LLM)-enhanced self-explanation techniques. Recognizing the barriers of language proficiency and cognitive overload, our approach leverages voice input and intelligent prompts to facilitate self-explanation, enabling students to articulate concepts in their own words. This process not only curbs the tendency to game educational systems but also supports students in organizing thoughts coherently, irrespective of their native language. The integration of LLMs provides adaptive feedback, simulating the critical presence of educators in virtual environments. Given the need for online videos and lectures again in the event of a pandemic, such systems can help encourage learning and provide students with feedback even when they are not sitting in front of educators in class.