Toxic Comments Classification in Social Networking

Abstract

        Today, users post a lot of comments on new portals, social networks, and forms. The majority of systems use some kind of automatic discovery of toxicity using machine learning models because it is impossible to manually moderate all of the comments. In this work, we performed a systematic review of the state-of-the-art in toxic comment classification using machine learning methods. We gathered information from 31 primary relevant studies. First, we looked into when and where the papers were written, as well as their maturity level. Every primary study's evaluation metric, used machine learning techniques, toxicity classes, and comment language were examined in our analysis. We conclude our work with a comprehensive list of gas currently being studied and suggestions for future research topics on the online toxic comment classification issue