Designing Human-Computer Collaborative Learning

Abstract

In this paper we describe the design of a system to support collaboration between a human learner and an artificial learner. The focus is on promoting the learner’s metacognitive skills by making strategic decisions explicit, by inducing reflection through criticism, and by fostering an active mode of learning. We outline the design of a system with which students can learn about electoral biases through designing electoral simulations. The implications for the machine learning component of the system, and for intelligent tutoring system architectures generally, are discussed.