Analysis of Noxious Comments using Machine Learning Algorithms

Noxious comments are impolite, rude, profane, or disrespectful online comments that can result in restricting people from expressing their thoughts and views, and to stop seeking others’ opinions because of the fear of harassment. Conversational toxicity is a snag that is making people uneasy and ultimately leading them to leave the discussion. This paper comprehensively examines the online abuse in text/comments and categorize the content into multi-label classification to study the toxicity as accurately as possible. Here, the impacts of Machine learning algorithms like Logistic regression, Naive Bayes, K-Nearest Neighbors (KNN), and Decision Tree have been studied and enforced to the facts to answer the troubles in textual categories, and the first-rate algorithm is chosen on the basis of metrics. Moving ahead, the work aims to achieve higher overall performance, and the levels of toxic content is studied for better accuracy to prevent its negative results which ultimately helps plenty of groups to inspect those issues.