Speech-based Psychological Crisis Assessment using LLMs

TL;DR

This paper introduces an LLM-based framework for automated psychological crisis assessment from speech, incorporating emotional cues and reasoning training to improve classification accuracy.

Contribution

It presents a novel paralinguistic injection method and reasoning-enhanced training strategy for better crisis level classification from spoken conversations.

Findings

Achieved macro F1-score of 0.802 and accuracy of 0.805

Enhanced emotional signal capture via non-verbal cue injection

Improved classification performance with reasoning and data augmentation

Abstract

Psychological support hotlines provide critical support for individuals experiencing mental health emergencies, yet current assessments largely rely on human operators whose judgments may vary with professional experience and are constrained by limited staffing resources. This paper proposes a large language model (LLM)-based framework for automated crisis level classification, a key indicator that supports many downstream tasks and improves the overall quality of hotline services. To better capture emotional signals in spoken conversations, we introduce a paralinguistic injection method that inserts identified non-verbal emotional cues into speech transcripts, enabling LLM-based reasoning to incorporate critical acoustic nuances. In addition, we propose a reasoning-enhanced training strategy that trains the model to generate diagnostic reasoning chains as an auxiliary task, which serves as a regulariser to improve classification performance. Combined with data augmentation, our final system achieves a macro F1-score of 0.802 and an accuracy of 0.805 on the three-class classification task under 5-fold cross-validation.