Towards Intuitive Interaction: Cognitive Architecture for Artificial Personality, Emotional Intelligence, and Cognitive Capabilities.

Robotic personality shows significant potential in enhancing Human-Robot Interaction. However, shortcomings at both theoretical and implementation levels lead to stereotypical and fragmented models of artificial personality. To overcome these limitations, this study aims to further validate the Conscientiousness, Extroversion, and Agreeableness (CEA) taxonomy for synthetic personalities based on the corresponding three traits of the Big Five model. In the proposed implementation, the robotic personality, directly inspired by human psychology, not only influences how actions are executed but also impacts the robot’s inner hedonic drives and, consequently, the action selection process. Our objective is to introduce a task- and platform-independent framework that drives agent behavior through the interplay between personality dynamics, emotional responses to others, memory encoding, anticipation of future actions, and associated hedonic experiences. To validate the effectiveness of the framework in generating perceivable personality traits and emotionally and cognitively intelligent behaviors, and to explore the impact of embodiment on social interaction, a dual experiment was conducted. This experiment involved participants engaging in dyadic conversations with either a digital human or a robot.