Evidence-Driven Marker Extraction for Social Media Suicide Risk Detection
This work addresses interpretability and efficiency challenges in LLM-based suicide risk detection for clinical applications, representing an incremental improvement over existing methods.
The paper tackled suicide risk detection from social media text by introducing Evidence-Driven LLM (ED-LLM), which uses multi-task learning to identify clinical markers and classify risk levels, achieving competitive classification performance and superior marker identification on the CLPsych datasets.
Early detection of suicide risk from social media text is crucial for timely intervention. While Large Language Models (LLMs) offer promising capabilities in this domain, challenges remain in terms of interpretability and computational efficiency. This paper introduces Evidence-Driven LLM (ED-LLM), a novel approach for clinical marker extraction and suicide risk classification. ED-LLM employs a multi-task learning framework, jointly training a Mistral-7B based model to identify clinical marker spans and classify suicide risk levels. This evidence-driven strategy enhances interpretability by explicitly highlighting textual evidence supporting risk assessments. Evaluated on the CLPsych datasets, ED-LLM demonstrates competitive performance in risk classification and superior capability in clinical marker span identification compared to baselines including fine-tuned LLMs, traditional machine learning, and prompt-based methods. The results highlight the effectiveness of multi-task learning for interpretable and efficient LLM-based suicide risk assessment, paving the way for clinically relevant applications.