Narrative-Based Robotic Assistant For Detecting and Alerting Major Depressive Episodes In Youth

This paper presents research on a technological environment supported by a nar-rative based robotic assistant designed to detect major depressive episodes. The study aims to validate how robotics can support young people with persistent de-pressive disorder and explore the potential to generate early alerts by analyzing both facial expressions captured by a camera and the emotional content of the re-sponses of the participants. The robotic assistant engages participants in a con-versational interaction and classifies their responses as positive, neutral, or nega-tive. These responses are aggregated, and when the number of negative responses exceeds a defined threshold, the system automatically triggers an alert notifica-tion.