Network-Based Hateful Communication Recognition System

Cyberbullying has emerged as a significant concern social media platforms, representing a global issue that negatively impacts both individuals and broader society. Detecting cyberbullying automatically is particularly challenging due to the nuanced and informal language commonly used online The short, casual nature of social media posts often leads vague or ambiguous expressions, complicating the interpretation of users intent. This complexity intensifies when dealing with uncertain context-dependent content. Although several existing methods aim to detect cyberbullying, they frequently struggle differentiate between the various subtle forms of hate speech due to overlapping meanings and inherent ambiguities. Moreover, many of these approaches fall short delivering high classificationaccuracy.