Classifying the Severity of Cyberbullying Incidents by Using a Hierarchical Squashing-Attention Network

Cyberbullying has become more prevalent in online social media platforms. Natural language processing and machine learning techniques have been employed to develop automatic cyberbullying detection models, which are only designed for binary classification tasks that can detect whether the text contains content. severity is a critical factor provide organizations with valuable information developing prevention strategies. This paper proposes hierarchical squashing-attention network (HSAN) classifying of incidents. Therefore, study aimed (1) establish Chinese-language dataset marked three ratings (slight, medium, serious) (2) new mechanism (SAM) HSAN according squashing function, uses vector length estimate weight attention. Experiments indicated SAM could sufficiently analyze sentences determine severity. The proposed model outperformed other machine-learning-based deep-learning-based models determining