Detection of Suicidal Twitter Posts

Abstract

As web data evolves, new technological challenges arise and one of the contributing factors to these challenges is the online social networks. Although they have some benefits, their negative impact on vulnerable users such as the spread of suicidal ideation is concerning. As such, it is vital to fine tune the approaches and techniques in order to understand the users and their context for early intervention. Therefore, in this study, we measured the impact of data manipulation and feature extraction, specifically using N-grams, on suicide-related social network text (tweets). We propose a diversified ensemble approach (multi-classifier fusion) to improve the detection of suicide-related text classification. Four machine classifiers were used for the fusion: Support Vector Machine, Random Forest, Naïve Bayes and Decision Tree. The results of our proposed approach have shown that the multi-classifier fusion has improved the detection of suicide-related text and, also, that Support Vector Machine has shown some promising results when dealing with multi-class datasets.