Adaptive Robot Discourse for Language Acquisition in Adulthood

Acquiring a second language in adulthood differs considerably from the approach taken at younger ages. Learning rates tend to decrease during adolescence, and socio-emotional characteristics, like motivation and expectations, take a different perspective for adults. In particular, acquiring communicative competence is a stronger objective for older learners, as an appropriate use of language in social contexts ensures a better community immersion and well-being. This skill is best attained through interactions with proficient speakers, but if this option is not available, social robots present a good alternative for this purpose. However, to obtain optimal results, a robot companion should adapt to the learner's proficiency level and motivation continuously to encourage speech production and increase flu-ency. Our work attempts to achieve this goal by developing an adaptive robot that modifies its spoken dialogue strategy, and visual feedback, to reflect a student's knowledge, proficiency and engagement levels in situated interactions for longterm learning.