The Japanese Society for Artificial Intelligence

Conversational agents represent an interactive computing tool as the solution to influencing people's opinions. In three settings of camera angles and facial expressions, 37 individuals interacted with virtual agents. Positive facial expressions from virtual agents and a near-distance camera angle successfully influenced participants' opinions of our primarily extroverted participants through empirical evaluations using ANOVA and the Big Five theory. There are undoubtedly various studies that are dispersed across numerous domains that could benefit from interactive agents. As a result, we came to understand the critical need of addressing a solid foundation for requirements to create an agent that aims to persuade people. To advance the field, this paper reveals the current experiment and future discussions for developing persuasive conversational agents.