Jessy
Ceha,
PhD
candidate
David
R.
Cheriton
School
of
Computer
Science
Conversational agents, or entities with some degree of 'intelligence' that have the capability of natural language discourse, are becoming common human-computer interfaces. One of the most heavily researched types of pedagogical conversational agents are based on tutoring scenarios. These types of agents can be distinguished by their role, either as teacher agents, co-learner agents, or teachable agents.
In this work we focus on teachable agents — systems that have the ability to be taught. Specifically, we examine the effects of a teachable social peer robot’s verbal expression of curiosity on students — how it is perceived, whether it can affect the emotional feeling and behavioural expression of curiosity in students themselves, and the impact on learning. Our findings indicate that verbally-expressed curiosity through on-task, topic-directed question-asking can reliably be recognized as curiosity in a teachable robot, and that curious robots can produce both emotional and behavioural curiosity contagion effects.