Classification of Depressive Symptoms Through Electrodermal Activity During Conversations with Virtual Humans: A Preliminary Study

Depressive disorder, frequently referred to as depression, is a common mental health condition that involves persistent feelings of sadness and loss of interest in daily activities. The use of objective techniques for the diagnosis and monitoring of this condition can significantly facilitate its management, in complement to traditional self-reports and clinical assessments. Based on these considerations, this study presents a preliminary investigation into the potential of electrodermal activity (EDA) in recognizing depressive symptoms. Additionally, to standardize the experimental protocol, emotional Virtual Humans are used to interact with subjects through conversations, thereby facilitating a controlled and repeatable environment. The aim is to develop an emotional recognition model to assess depressive symptoms, supporting clinicians and sufferers in diagnosing and monitoring depression. The experimental findings show promising results, suggesting that EDA holds the potential to facilitate a deeper understanding of depressive symptoms. This can lead to improved diagnostic accuracy and personalized treatment approaches for depression management.