Toward Stopping Incel Rebellion: Detecting Incels in Social Media Using Sentiment Analysis

Abstract

Incels, which stand for involuntary celibates, refer to online community members who identify themselves as individuals that women are not attracted to them. They are usually involved in misogyny and hateful conversations on social networks, leading to several terrorist attacks in recent years, also known as incel rebellion. In order to stop terrorist acts like this, the first step is to detect incels members in social networks. To this end, user-generated data can give us insights. In previous attempts to identifying incels in social media, users’ likes and fuzzy likes data were considered. However, another piece of information that can be helpful to identify such social network members is users’ comments. In this study, for the first time, we have considered users’ comments to identify incels in the social networks. Accordingly, an algorithm using sentiment analysis was proposed. Study results show that by implementing the proposed method on social media users’ comments, incel members can be identified in social networks with an accuracy of 78.8%, which outperforms the previous work in this field by 10.05%.