Multi-Task Learning for Mental Health using Social Media Text

TL;DR

This paper presents a multi-task deep learning approach to predict suicide risk and mental health issues from social media text, demonstrating improved accuracy over single-task models especially with limited data.

Contribution

It introduces a multi-task learning framework that jointly models mental health conditions and gender, showing significant performance gains over traditional single-task methods.

Findings

Best MTL model achieves AUC > 0.8 for suicide attempt prediction

Multi-task learning improves performance on limited data scenarios

Joint modeling of multiple conditions enhances prediction accuracy

Abstract

We introduce initial groundwork for estimating suicide risk and mental health in a deep learning framework. By modeling multiple conditions, the system learns to make predictions about suicide risk and mental health at a low false positive rate. Conditions are modeled as tasks in a multi-task learning (MTL) framework, with gender prediction as an additional auxiliary task. We demonstrate the effectiveness of multi-task learning by comparison to a well-tuned single-task baseline with the same number of parameters. Our best MTL model predicts potential suicide attempt, as well as the presence of atypical mental health, with AUC > 0.8. We also find additional large improvements using multi-task learning on mental health tasks with limited training data.