ABSTRACT
AutoTutor is a complex cognitive system that simulates a human tutor, or an ideal tutor, by holding a conversation with the learner in natural language. AutoTutor qualifies as a cognitive system in two fundamental ways (see Forsythe & Xavier, chap. 1, this volume). First, its core architecture incorporates a humanlike model of knowledge, discourse, and cognitive processes. Second, AutoTutor recruits these psychological mechanisms to engage the learner cognitively and thereby promote learning. Therefore, AutoTutor both simulates cognitive mechanisms of human tutors and facilitates cognitive activities of human learners.
