» Articles » PMID: 36231935

Detecting and Analyzing Suicidal Ideation on Social Media Using Deep Learning and Machine Learning Models

Overview
Publisher MDPI
Date 2022 Oct 14
PMID 36231935
Authors
Affiliations
Soon will be listed here.
Abstract

Individuals who suffer from suicidal ideation frequently express their views and ideas on social media. Thus, several studies found that people who are contemplating suicide can be identified by analyzing social media posts. However, finding and comprehending patterns of suicidal ideation represent a challenging task. Therefore, it is essential to develop a machine learning system for automated early detection of suicidal ideation or any abrupt changes in a user's behavior by analyzing his or her posts on social media. In this paper, we propose a methodology based on experimental research for building a suicidal ideation detection system using publicly available Reddit datasets, word-embedding approaches, such as TF-IDF and Word2Vec, for text representation, and hybrid deep learning and machine learning algorithms for classification. A convolutional neural network and Bidirectional long short-term memory (CNN-BiLSTM) model and the machine learning XGBoost model were used to classify social posts as suicidal or non-suicidal using textual and LIWC-22-based features by conducting two experiments. To assess the models' performance, we used the standard metrics of accuracy, precision, recall, and F1-scores. A comparison of the test results showed that when using textual features, the CNN-BiLSTM model outperformed the XGBoost model, achieving 95% suicidal ideation detection accuracy, compared with the latter's 91.5% accuracy. Conversely, when using LIWC features, XGBoost showed better performance than CNN-BiLSTM.

Citing Articles

Risk-based evaluation of machine learning-based classification methods used for medical devices.

Haimerl M, Reich C BMC Med Inform Decis Mak. 2025; 25(1):126.

PMID: 40069689 PMC: 11895222. DOI: 10.1186/s12911-025-02909-9.


Mining Suicidal Ideation in Chinese Social Media: A Dual-Channel Deep Learning Model with Information Gain Optimization.

Meng X, Cui X, Zhang Y, Wang S, Wang C, Li M Entropy (Basel). 2025; 27(2).

PMID: 40003113 PMC: 11854197. DOI: 10.3390/e27020116.


Suicide Risk Assessment on Social Media with Semi-Supervised Learning.

Lovitt M, Ma H, Wang S, Peng Y Proc IEEE Int Conf Big Data. 2025; 2024:8541-8549.

PMID: 39896202 PMC: 11786971. DOI: 10.1109/bigdata62323.2024.10825422.


An Explainable Artificial Intelligence Text Classifier for Suicidality Prediction in Youth Crisis Text Line Users: Development and Validation Study.

Thomas J, Lucht A, Segler J, Wundrack R, Miche M, Lieb R JMIR Public Health Surveill. 2025; 11:e63809.

PMID: 39879608 PMC: 11822322. DOI: 10.2196/63809.


The steps that young people and suicide prevention professionals think the social media industry and policymakers should take to improve online safety. A nested cross-sectional study within a Delphi consensus approach.

Robinson J, Thorn P, McKay S, Richards H, Battersby-Coulter R, Lamblin M Front Child Adolesc Psychiatry. 2025; 2:1274263.

PMID: 39839592 PMC: 11748789. DOI: 10.3389/frcha.2023.1274263.


References
1.
Alkahtani H, Aldhyani T . Artificial Intelligence Algorithms for Malware Detection in Android-Operated Mobile Devices. Sensors (Basel). 2022; 22(6). PMC: 8954874. DOI: 10.3390/s22062268. View

2.
Braithwaite S, Giraud-Carrier C, West J, Barnes M, Hanson C . Validating Machine Learning Algorithms for Twitter Data Against Established Measures of Suicidality. JMIR Ment Health. 2016; 3(2):e21. PMC: 4886102. DOI: 10.2196/mental.4822. View

3.
Al-Adhaileh M, Aldhyani T, Alghamdi A . Online Troll Reviewer Detection Using Deep Learning Techniques. Appl Bionics Biomech. 2022; 2022:4637594. PMC: 9213121. DOI: 10.1155/2022/4637594. View

4.
Klonsky E, May A . Differentiating suicide attempters from suicide ideators: a critical frontier for suicidology research. Suicide Life Threat Behav. 2013; 44(1):1-5. DOI: 10.1111/sltb.12068. View

5.
Sueki H . The association of suicide-related Twitter use with suicidal behaviour: a cross-sectional study of young internet users in Japan. J Affect Disord. 2014; 170:155-60. DOI: 10.1016/j.jad.2014.08.047. View