Measurements and Interventions to Improve Student Engagement Through Facial Expression Recognition

Will Lee, Danielle Allessio, William Rebelsky, Sai Satish Gattupalli, Hao Yu, Ivon Arroyo, Margrit Betke, Sarah Adel Bargal, Tom Murray, Frank Sylvia, Beverly P. Woolf

Abstract: A major challenge for online learning systems is supporting students’ engagement. Online systems are sometimes boring, repetitive, and unappealing; external distractions often lead to off-task behavior, and a decline in learning. Student engagement and emotion are also tightly correlated with learning gains because emotion drives attention and attention drives learning. In response to this challenge, we present an exploratory study using computer vision techniques and several engagement strategies to help students re-engage once their attention wanders and their head turns away, within the context of an intelligent tutoring system for mathematics. Initial results indicate that students exposed to our re-engagement strategies were more confident and more persistent. They also responded positively to our avatar or learning companion. We discuss design implications and future work.