Contrastive Learning and Large Language Models for Depression Detection from Social Media

Abstract

Depression is a global mental health concern, often expressed subtly through social media platforms like Reddit and Twitter. This study explores the application of contrastive learning with large language models (LLMs) for depression detection, utilizing annotated data from the DepressionEmo dataset.