Consensus-Based Ensemble Model for Arabic Cyberbullying Detection

Due to the proliferation of internet-enabled smartphones, many people, particularly young people in Arabic society, have widely adopted social media platforms as a primary means communication, interaction and friendship making. The technological advances smartphones communication enabled keep touch form huge networks from all over world. However, such expose cyberbullying offensive content that puts their safety emotional well-being at serious risk. Although, solutions been proposed automatically detect cyberbullying, most existing designed for English speaking consumers. morphologically rich languages-such language-lead data sparsity problems. Thus, render developed another language are ineffective once applied content. To this end, study focuses on improving efficacy detection models by designing developing Consensus-based Ensemble Cyberbullying Detection Model. A diverse set heterogeneous classifiers traditional machine deep learning technique trained using labeled dataset collected five different platforms. outputs selected combined consensus-based decision-making which F1-Score each classifier was used rank classifiers. Then, Sigmoid function, can reproduce human-like decision making, is infer final decision. outcomes show model comparing other studied overall improvement gained reaches 1.3% with best classifier. Besides its effectiveness content, be generalized improve languages.