This research focuses on the identification of cybercrimes occurring on the social media platform Twitter, with the objective of mitigating the rising prevalence and expansion of online security risks. The suggested system employs sentiment analysis methodologies to detect instances of harassment or threats within social media content. Machine learning algorithms, namely Naive Bayes, Logistic Regression, SVM, and CNN are compared to determine their effectiveness in detecting cyberbullying activity. The goal is to achieve the highest accuracy in identifying threats and alerting users to potential criminal activity in real-time. By implementing such a system, users can have a more secure social media experience by staying informed and vigilant about potential risks in both the virtual and real worlds.

bitch, jihad, terrorism, defamation, blackmail, radicalization, fool.