Prediction Of Hate Speech Classification Using Secured Supervised Machine Learning With Nlp

Abstract

The daily lives of many individuals have been profoundly impacted by digital media. Hate speech refers to a story that is meant to distract or deceive the audience. Due to a number of factors, including the rise of online social networks during the recent years, hate speech has increased in frequency in the online world. Online social network users can easily be impacted by this hatred speech. Hate speech has become an issue for society, sometimes spreading more quickly than the truth. All of this hate speech is undetectable to a person. Therefore, a machine learning algorithm that can recognize hate speech automatically is required. Machine learning models are developed using algorithms to determine whether a speech is hate speech, hurtful speech, or neither. When compared to the other algorithms, the Gradient Boosting Algorithm produces the greatest accuracy. Therefore, the Gradient Boosting Algorithm is used for project launch. The Kaggle dataset used for this project to classify hate speech contains features like count, hate speech, offensive, neither hate nor offensive, class, and tweet.