Intelligent tutoring systems and conversational agents

This article reviews advances in computerized intelligent tutoring systems, particularly those with conversational agents that simulate human tutors, peers, and mentors. Empirical research on learning gains, engagement, and learner impressions have confirmed the promise of these digital technologies. The article identifies major components of these systems and presents example systems that cover different subject matters, skills, learner populations, and pedagogical functions. We also point out major challenges that need to be addressed for scaling up these systems to millions of learners throughout the globe.