Detecting Virtual Harassment in Social Media Using Machine Learning

The escalating prevalence of cyberbullying stands as a significant challenge in our digital era, causing severe repercussions on individuals' mental health, well-being, and professional integrity. This paper introduces an innovative approach to combat this online menace. We present a cyberbullying detection model based on machine learning principles. Our method involves curating a dataset comprising a variety of messages pre-labeled into multiple harassment categories. By leveraging sophisticated machine learning techniques, we extract distinctive features from these messages, enabling precise identification of cyberbullying instances. The results obtained underscore the high accuracy of our detection model, highlighting its undeniable effectiveness in combating online cyberbullying. This research thus offers a significant contribution by presenting an innovative method to detect and address this concerning phenomenon on digital platforms.