Designing effective behaviors for educational embodied agents

Abstract

While various forms of online and distance learning have opened up opportunities for people to gain knowledge independent of where they are, embodied interaction with teachers or learning partners supports the social and cognitive aspects of the learning processes. To achieve effective and efficient teaching, teachers use nonverbal behaviors to communicate the focus of their attention and highlight important concepts. With the advance of technology, embodied agents (i.e., robots) keep showing promises of supporting human activities. The goal of my dissertation is to design effective nonverbal behaviors, with a concentration on gaze, hand gestures, and head movements, for embodied agents as teachers or learning partners to support people in learning. To this end, my research focuses on understanding how and what nonverbal behaviors contribute to effective learning, implementing behavioral patterns from this understanding on embodied agents, and evaluating the effectiveness of the embodied agents in various educational contexts.