Beyond Risk Prediction: Evidence Grounding and Psychosocial Factor Verification for Explainable Suicide Risk Assessment
Summary
这篇论文提出一个用于可解释自杀风险评估的框架，包含风险评估、证据定位和心理社会因素识别三个模块，并在 IEEE BigData 2026 Explainable Suicide Risk Detection Challenge 中报告了相应的 F1 结果。
View Cached Full Text
Cached at: 10/08/26, 12:38 PM
# Beyond Risk Prediction: Evidence Grounding and Psychosocial Factor Verification for Explainable Suicide Risk Assessment Source: [https://arxiv.org/abs/2610.08842](https://arxiv.org/abs/2610.08842) [View PDF](https://arxiv.org/pdf/2610.08842)[HTML \(experimental\)](https://arxiv.org/html/2610.08842v1) > Abstract:Identifying suicide risk from social networking services \(SNS\) posts is important for detecting suicide\-related signals in online environments\. However, risk classification alone provides limited insight into the textual evidence and psychosocial factors behind a prediction\. Based on the IEEE BigData 2026 Explainable Suicide Risk Detection Challenge, this study presents a framework consisting of Risk Assessment, Evidence Grounding, and Factor Identification\. Risk Assessment uses length\-based routing to accommodate posts of different lengths\. Evidence Grounding identifies supporting phrases and uses a Risk\-Evidence constraint to maintain consistency with the Risk prediction\. For Factor Identification, two verifiers are used\. The Taxonomy Verifier focuses on factor semantics, whereas the Evidence\-Aware Verifier uses factor\-specific lexical\-semantic cues to select informative positive training units\. Their prediction probabilities are combined to produce the final factor predictions\. The three tasks are evaluated using task\-specific F1 score measures\. Risk Assessment achieved a Weighted F1 of 0\.8088, Evidence Grounding achieved a test Macro row F1 of 0\.7605, and Factor Identification achieved a Macro F1 of 0\.5562\. The results show that the framework can provide risk predictions, along with supporting textual evidence and fine\-grained information on psychosocial factors\. Overall, the proposed framework extends suicide\-risk assessment beyond risk\-level prediction and provides a more interpretable analysis of SNS posts\. ## Submission history From: Tianle Hu \[[view email](https://arxiv.org/show-email/9b489cb6/2610.08842)\] **\[v1\]**Wed, 30 Sep 2026 12:07:01 UTC \(436 KB\)
Similar Articles
Evidence-Anchored Overdose Death Prediction from Death Investigation Reports
A research paper presenting an automated framework that classifies suspected overdose deaths from death investigation reports using expert-guided explanations, showing that a pipeline model with input-grounded rationales outperforms self-rationalizing models and a deployed rule-based system by 15%.
Building Trustworthy Mental Health Benchmarks on Bluesky: A Validation-Aware Weak-Supervision Framework
This paper introduces a validation-aware weak-supervision framework for building and evaluating suicidal ideation and mental health disclosure benchmarks on the decentralized social media platform Bluesky, leveraging AI models like Llama-3-8B and transformer-based classifiers to analyze performance and label validity.
Ensemble Feature Selection and Harris Hawks Optimization for Explainable Mental Health Risk Prediction in Female Sex Workers
This paper proposes a hybrid predictive model combining ensemble feature selection (ANOVA and mutual information) with Harris Hawks optimization-tuned logistic regression for explainable mental health risk prediction in female sex workers, achieving 95.78% accuracy.
Estimating suicide risk from text
MIT researchers developed a language-processing tool that uses text analysis to estimate suicide risk, published in the Journal of Psychopathology and Clinical Science with potential applications in crisis counseling.
DistressBench: Evaluating Multi-Dimensional Crisis Support in Large Language Models
DistressBench 提出一个由临床医生编写的 718 段对话基准，用于评估大模型在自杀与自残危机情境下的多维度支持能力，发现前沿模型平均只能满足 71.4% 的必要照护标准，且常在识别危机后未能引导用户寻求人工帮助。