Search

This research presents a novel framework for distinguishing between actual and non-suicidal ideation in social media interactions using an ensemble technique. The prompt identification of sentiments on social networking platforms is crucial for timely intervention serving as a key tactic in suicide prevention efforts. However, conventional AI models often mask their decision-making processes primarily designed for classification purposes. Our methodology, along with an updated ensemble method, bridges the gap between Explainable AI and leverages a variety of machine learning algorithms to improve predictive accuracy. By leveraging Explainable AI's interpretability to analyze the features, the model elucidates the reasoning behind its classifications leading to a comprehension of hidden patterns associated with suicidal ideations. Our system is compared to cutting-edge methods on several social media datasets using experimental evaluations, demonstrating that it is superior, since it detects suicidal content more accurately than others. Consequently, this study presents a more reliable and interpretable strategy (F1-score for suicidal = 95.5% and Non-Suicidal = 99%), for monitoring and intervening in suicide-related online discussions., Competing Interests: Declarations. Competing interests: The authors declare no competing interests. Informed consent: All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5). Additional informed consent was obtained from all patients for whom identifying information is included in this article. Human and animal rights: This study did not involve any experimental research on humans or animals; hence, an approval from an ethics committee was not applicable in this regard. The data collected from the online forums are publicly available data, and no personally identifiable information of the forum users were collected or used for this study., (© 2025. The Author(s).)