Predicting Harmful Web Pages Based on Suicide-related Textual Analysis using Machine Learning Algorithms

Abstract

And detecting suicidal people remains a difficult task. As the usage of social media has grown, we've seen people openly discuss their suicide plans or attempts on these platforms. Suicide prevention is addressed in this article by identifying suicidal profiles on social networks. First, we examine online profiles and extract a variety of information, such as account features connected to the profile and features relevant to the social media data. Second, we present our technique for detecting suicidal profiles using Twitter data, which is based on machine learning algorithms. Then, as a profile data set, we employ a data set of people who have previously committed suicide. The efficiency of our technique in terms of memory and precision in detecting suicidal characteristics is supported by experimental data. Finally, we demonstrate the detection of suicidal characteristics using a Java-based prototype of our work.