Assessing the Efficiency of Machine Learning Algorithms in Detecting Cyberbullying on Twitter

Abstract

Cyberbullying represents one of the unwanted behaviors which has gotten more and more common since the use of social media, particularly Twitter, has expanded quickly. Cyberbullying can have serious repercussions for its victims. Due to large volumes of data and the intricate structure of online interactions, detecting cyberbullying in real-time is an exhausting task. The purpose of this research proposal is to utilize machine learning techniques that will enhance precision and recall in recognizing instances of cyberbullying on Twitter. In the study, Twitter data is evaluated, trends and patterns are identified, and a wide range of cooperative behaviors are thoroughly assessed. The creation of a framework for collaboration, empirical analysis of patterns of collaboration, evaluation of performance indicators, and improvement of real-time detection capabilities are the main goals.