Collaborative robot assisted language learning

Abstract:
Social robots have entered the education sector and showed great
promises. Robot assisted language learning, in particular, has gained a
considerable momentum in the last couple years, as social robots can
offer a more natural way to practice new languages than apps or computer
programs. However, when target activity is a practice conversation, new
challenges are introduce in the robot-learner interaction. Some of them
come from the technical limitations of having a non-native speaker,
while others are related to the requirement of sustaining an interaction
that also supports learning.

Our work explores this robot-led conversation practice for multi-party
interactions, where we attempt to adapt the verbal and non-verbal
behavior of the social robot to promote collaborative learning. In this
seminar, I will cover the analysis to detect learners’ uncertainty or
confusion during the practice conversation, the effect of adapting
non-verbal behaviors to influence group dynamics in a language learning
game, and our current efforts to explore human-robot collaboration to
improve learning.