Text Analysis for Depression Detection: Mental Health Digital Transformation

Abstract

Depression is a pervasive mental health disorder affecting millions globally. The rise of social networks and their digital footprint provides a unique avenue to leverage AI for early identification of users who may be suffering. We built upon BERT for feature extraction from individual user posts, followed by a Convolutional Neural Network for classification. Since the pre-trained BERT model may not effectively capture social media language, we propose an approach to pre-train BERT on Reddit data before integrating it into the BERT+CNN architecture.