Depression Risk Assessment in Social Media via Large Language Models

ArXi:2604.19887v1 Announce Type: new Depression is one of the most prevalent and debilitating mental health conditions worldwide, frequently underdiagnosed and undertreated. The proliferation of social media platforms provides a rich source of naturalistic linguistic signals for the automated monitoring of psychological well-being. In this work, we propose a system based on Large Language Models (LLMs) for depression risk assessment in Reddit posts, through multi-label classification of eight depression-associated emotions and the computation of a weighted severity index.