Scott W. Greenwald

Abstract

The planned thesis presents a vision of the future of learning, where learners explore environments, physical and virtual, in a curiosity-driven or intrinsically motivated way, and receive contextual information from a companion facilitator or teacher. Learners are instrumented with sensors that convey their cognitive and attentional state to the companion, who can then accurately judge what is interesting or relevant, and when is a good moment to jump in. I provide a broad definition of the possible types of sensor input as well as the modalities of intervention, and then present a specific proof-of-concept system that uses gaze behavior as a means of communication between the learner and a human companion.