EAGER: Multi-Modal Communication for Teachable Robots

点击翻译按钮获取中文摘要

Peer tutoring, in which one student teaches material to another student, has been proven to increase learning, especially for the student who is taking on the role of teacher. This is because detecting and correcting mistakes often requires the student teacher to reflect on their own knowledge to develop explanations, reinforcing their knowledge of the subject matter. The effectiveness of such peer tutoring, however, depends on how learners behave -- if they pick up the subject too quickly, the student teacher may benefit less; if they pick it up too slowly, the teacher may get frustrated or confused themselves. An alternative is to have the learner pretend not to know the subject and help guide the teacher by "back leading" -- asking the right questions, making occasional "mistakes", etc. To this end, the goal of this research is to develop a social robot that plays the role of learner, guiding the student teacher using both verbal and non-verbal behaviors, modeling both the cognitive and emotional state of the student, and providing feedback and encouragement.The project focuses on developing multi-modal communication for a social robot that provides cognitive and emotional support of students who are trying to teach the robot multi-step mathematical problem solving. The student and robot collaborate around a "smart" whiteboard that forms the joint focal point for learning. Concepts from Theory of Mind are used to infer the affective state of the student from their multi-modal behaviors and to provide support and encouragement, with the goal of facilitating the student persevering in the task and developing meta-cognitive skills to teach more effectively. Concepts from Common Ground Theory are used to generate robot actions, such as asking questions or making deliberate mistakes, to help guide the student teacher to better explain the concept, through an interactive presentation/acceptance process. As part of this project, a robot ambassador is being developed that can explain robotics and STEM to middle-school students.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.