Classification of Cyber Hate Speech from Social Networks using Machine Learning

Social networking sites like Twitter, Facebook, and others play a big part in social network analysis in the present Web 2.0 environment. Hate speech, which is abusive communication that singles out certain group traits like gender, religion, or race in order to incite violence, is growing in importance as social media platforms gain popularity. Hate speech on the internet is a relatively new problem in our society that is steadily growing by exploiting the flaws in the platforms that distinguish most social networking sites. The main source of this occurrence is offensive remarks, whether delivered during user contact or in the form of an uploaded multimedia context. Hateful and toxic content created by a portion of social media users is a growing phenomenon that has prompted researchers to devote significant resources to the difficult task of identifying hateful content. Some of the popular methods are the Support Vector Machine, the Logistic Regression Model, and Decision Trees. These strategies, however, frequently come under the umbrella of discriminative learning, which seeks to distinguish one class from others while taking into consideration the real world. First, this study has reviewed social networks and social network analysis. Second, the necessity of detecting hate speech and the distinction between it and abusive content are covered. Thirdly, several machine learning algorithms are being contrasted. The experimental findings demonstrate that the suggested strategy consistently outperforms the alternatives.