Mutual Human-Computer Tutoring Using Agents

Abstract

Learning platforms provide educational tools and strategies to impart the necessary knowledge, skills and attitudes to learners. Effectiveness and the ability to deliver tangible benefits to learners largely depend on how they are designed, delivered and assessed. Smart tutoring systems are a solution to overcome the difficulties of learners by allowing them to have personalized learning and follow-up. This article presents a critique of the effectiveness of these platforms and proposes a generic architecture independent of the learning subject and which offers the learner Machine / Human monitoring.