Multi Label Toxic Comment Classification using Machine Learning Algorithms

Abstract

Toxic comments are the comments found in the online forums that are rude, offensive, or unfair and usually cause many users to exit the conversation. The threat of bullying and abuse on the internet obstructs the free exchange of ideas by limiting people’s opposing viewpoints. Most of the Websites fail to successfully facilitate healthy conversations, leading them to either restrict or disable user comments entirely. This paper would explore the scope of online abuse and categorize them into different labels to assess the toxicity as accurately as possible using machine learning algorithms.