EMPATHIES: HUMAN AND DIGITAL BODIES An interdisciplinary approach to enhancing human–conversational agent interaction

The evolution of embodied conversational agents (ECAs) opens new perspectives in human-computer communication, combining verbal and non-verbal language to foster more empathetic interaction. The following study adopts an interdisciplinary approach to develop an algorithm capable of recognizing gestures of greeting, agreement, and disagreement. The research is based on data collection and the use of emotional stimuli, such as avatars’ facial expressions and music, to elicit nonverbal responses useful for training the system. Although ECAs do not experience human emotions, they can provide rational empathy, optimize the user experience and enabling smoother, more natural interactions. Technological advancements give designers new tools for creating effective interactions.