A Two-Stage Selective Escalation Framework for Robust Suicide Risk Detection on Social Media: Machine Learning Study (Preprint)

Read full text Summarise with VIRAT
Authors: Yukai Song, Pengfei Zhou, César G. Escobar-Viera, Candice L. Biernesser, Wei Huang, Jingtong Hu
Cited by: 0
References: 21

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

BACKGROUND Suicide rates have risen worldwide in recent years, underscoring the urgent need for proactive and scalable prevention strategies. Social media has emerged as a valuable source for early screening, as many at-risk individuals who avoid formal help due to stigma express distress online. However, current suicide risk detection systems face a critical limitation: while explicit suicidal expressions are relatively easy to identify, implicit suicidal ideation conveyed through metaphor, sarcasm, or subtle emotional cues remains difficult to detect and constitutes a major source of system failure in real-world deployment. OBJECTIVE This study aims to develop a robust and computationally efficient suicide risk detection framework that can reliably handle both explicit and implicit expressions while controlling inference cost and improving practical deployability for mental health screening. METHODS We propose a robust and efficient two-stage framework that explicitly accounts for heterogeneous signal strength through selective escalation. In Stage 1, a lightweight Bidirectional Encoder Representations from Transformers (BERT) classifier rapidly resolves high-confidence explicit cases. In Stage 2, only ambiguous inputs are escalated for further analysis through two complementary mechanisms: (i) a multi-perspective large language model (LLM) voting strategy designed to improve recall on implicit suicide risk, and (ii) a feature-based machine learning ensemble guided by psychologically grounded indicators extracted via prompt-engineered LLMs, enabling efficient inference and enhanced interpretability. RESULTS We evaluate the proposed framework on two complementary datasets representing distinct risk profiles: an explicit-dominant Reddit dataset and an implicit-only DeepSuiMind dataset. The framework achieves an F1 score of 98.0% on the Reddit benchmark and up to 99.94% recall on the implicit-only DeepSuiMind benchmark (positive-only; precision is not defined). It reduces the cross-domain performance gap to approximately 1.2-1.4% in recall and 1.7-2.2% in F1, and significantly lowers LLM invocation costs compared with single-model baselines. CONCLUSIONS These results demonstrate that system-level design with selective escalation offers a practical and effective pathway for suicide risk detection systems that are both robust to implicit expressions and efficient enough for real-world mental health screening applications. CLINICALTRIAL N/A

Fields of Study & Subjects

Mental Health via Writing Suicide and Self-Harm Studies Hate Speech and Cyberbullying Detection

Publication details

Publication date
19 March 2026
Language
EN
Document type
Preprint