Identify High-Risk Suicidal Posts and Psychological Risk Factors on Social Media Using a Two-Stage Deep Learning Model
Abstract: Our study aims to utilize psychological risk factors to detect articles on social media that are at high risk for suicidal content. We propose a two-stage model structure: the first stage labels each sentence in an article with risk factors, and the second stage uses this information as features to predict the crisis level of the article. Our models were trained using a dataset that we developed, which consists of social media posts from Dcard. These posts were labeled by psychological professionals and will be publicly released. Our approach achieved an accuracy and F1-score of 0.96 in classifying high-crisis-level articles. Our research facilitates the automatic detection of high-crisis-level articles for further analysis of risk factors, enhancing interdisciplinary collaboration between natural language processing, deep learning, and psychology.
Paper Type: Long
Research Area: Computational Social Science and Cultural Analytics
Research Area Keywords: psycological NLP, emotion detection and analysis, social media, suicide prevention
Contribution Types: Model analysis & interpretability, NLP engineering experiment, Data resources
Languages Studied: Chinese
Submission Number: 4714
Loading