Robots in the classroom: Learning to be a Good Tutor
Emmanuel Senft, Séverin Lemaignan, Madeleine E. Bartlett, Paul E. Baxter, Tony Belpaeme · Ghent University Academic Bibliography (Ghent University) · 2018
To broaden the adoption and be more inclusive, robotic tutors need to tailor theirbehaviours to their audience. Traditional approaches, such as Bayesian KnowledgeTracing, try to adapt the content of lessons or the difficulty of tasks to the currentestimated knowledge of the student. However, these variations only happen in a limiteddomain, predefined in advance, and are not able to tackle unexpected variation in astudent's behaviours. We argue that robot adaptation needs to go beyond variations inpreprogrammed behaviours and that robots should in effect learn online how to becomebetter tutors. A study is currently being carried out to evaluate how human supervisioncan teach a robot to support child learning during an educational game using oneimplementation of this approach.