FinAIPhys: A Data Collection Platform for Future AI-Based Adaptive Learning in Mechanics

Abstract

One recurring challenge in physics education is that students arrive at class with deeply rooted misconceptions, which place a heavy challenge on both them and their teachers to correct. Adaptive, data-driven digital learning environments offer a promising tool for supporting personalized learning that promotes conceptual change. This article presents the data collection algorithm of a digital learning environment called FinAIPhys. The system serves a dual purpose. It collects large amounts of multidimensionally labelled student responses intended to support the future development of an AI-based recommendation system, and it supports student users in preparing for their physics final exams.