An Automatic Method to Prevent and Classify Cyberbullying Incidents using Machine Learning Approach

Abstract

The technological advancements and the increasing popularity of social networking platforms, the sharing of personal information among online users has become widespread. This sharing occurs effortlessly through various devices such as computers and mobile phones. Cyberbullying can manifest through SMS, text messages, and various applications, as well as online platforms like social media and forums, where individuals can view, engage with, or distribute content.The project offers a comprehensive understanding of Cyberbullying incidents and their corresponding offences combining a series of approaches reported in relevant Work. The implementation provides the opportunity to systematically combine various element or Cyberbullying characteristics. Additionally, a comprehensive list of Cyberbullying-related offences is put forward. The offenses are ordered in a Deep Neural Network classification system based on specific criteria to assist in better classification and correlation of their respective incidents. This enables a thorough understanding of the repeating and underlying criminal activities. This study focuses on classifying user posts and image content into bullying or non bullying through reputation score