Published Paper Details:

The evolution and growth of social networking and modern web technology have made an individual's online presence permanent. People frequently express their thoughts, ideas, and emotions through social networking links, with the most popular activity being the discussion of everyday events, which may include private or public conversations. Bullying that involves technology is referred to as cyber-bullying. Bullying attacks target teenagers and young people who spent lot of time on social networking sites. The rise of social media, especially Twitter, has caused confusion about the meaning of free expression, which has given rise to a number of worries. Cyber-bullying is one of these issues, a severe global problem that has an impact on both people and societies. There have been numerous reported attempts to intervene, prevent, or lessen cyber-bullying; however, these efforts are unworkable since they depend on the victims' interactions. Victims of this behavior may experience hopelessness and other potentially fatal issues. It is necessary to create monitoring and detection procedures for potentially hazardous Internet behavior. By using machine learning, we can create algorithms to automatically identify cyber-bullying content and recognize language patterns used by bullies and their victims. Therefore, it is crucial to identify cyber-bullying without the victims' participation. Additionally, numerous machine learning classifiers were applied, including the K-nearest neighbor technique, linear regression, decision trees, and the Support Vector Machine classifiers, Random Forest, Naive Bayes, and AdaBoost. The most precise supervised learning algorithm was used to identify communications that contained cyber-bullying.