Publication

Abstract : Cyberbullying is a severe issue that impacts teens and adults alike. Errors like hopelessness and suicide have resulted from it. The demand for material on social media platforms to be regulated is growing. In the work that follows, machine learning and natural language processing are used to construct a model to identify cyberbullying in text data. The program leverages data from two different types of cyberbullying: personal attacks on Wikipedia forums and tweets containing hate speech. To determine the optimal approach, three feature extraction techniques and four classifiers are examined. The method produces more than 80% accurate data from Wikipedia and more than 90% correct data from Twitter.