Abstract: Social Robots have been shown to promote student learning and engagement through interactions with students as peer-like 'learning companions'. In this talk, I will discuss research projects introducing lifelong and transferrable models for adaptive student personalization, evaluated within a suite of tablet-based games designed to promote students' language and literacy skills through interactive play with social robots. I will also discuss work on personalizing to student affective cues, using dynamic estimates of student facial expressions and body posture to adapt a robot's content and behavior in a tutoring interaction to promote learning and engagement. By extending personalized modeling to 'long-term' and 'multi-task' personalized interactions, and expanding the modalities along which personalization occurs, this work contributes both algorithmic and human-centered insights for the future of educational human-robot interactions.

Short Bio: Sam Spaulding is a final year PhD student in the Personal Robots Group at the MIT Media Lab, advised by Cynthia Breazeal. His thesis research focuses on developing and evaluating the underlying technologies for social robot learning companions to support early-language and literacy skills, synthesizing insights from machine learning, affective and educational sciences, and interactive media. His research has been published and presented at international conferences including AAAI, HRI, AAMAS, and CogSci and has been recognized by an NSF Graduate Research Fellowship, a Human-Robot Interaction Pioneers award, and research grants from the Mellon Fund and Sigma Xi scientific society.