Open Access

ARTICLE

Enhancing Multi-Class Cyberbullying Classification with Hybrid Feature Extraction and Transformer-Based Models

Suliman Mohamed Fati^1,*, Mohammed A. Mahdi², Mohamed A.G. Hazber², Shahanawaj Ahamad³, Sawsan A. Saad⁴, Mohammed Gamal Ragab⁵, Mohammed Al-Shalabi²

1 Information Systems Department, College of Computer and Information Sciences, Prince Sultan University, Riyadh, 11586, Saudi Arabia
2 Information and Computer Science Department, College of Computer Science and Engineering, University of Ha’il, Ha’il, 55476, Saudi Arabia
3 Software Engineering Department, College of Computer Science and Engineering, University of Ha’il, Ha’il, 55476, Saudi Arabia
4 Computer Engineering Department, College of Computer Science and Engineering, University of Ha’il, Ha’il, 55476, Saudi Arabia
5 Department of Computer and Information Sciences, Universiti Teknologi Petronas, Seri Iskandar, 32610, Malaysia

* Corresponding Author: Suliman Mohamed Fati. Email:

(This article belongs to the Special Issue: Emerging Artificial Intelligence Technologies and Applications)

Computer Modeling in Engineering & Sciences 2025, 143(2), 2109-2131. https://doi.org/10.32604/cmes.2025.063092

Received 05 January 2025; Accepted 08 April 2025; Issue published 30 May 2025

Abstract

Cyberbullying on social media poses significant psychological risks, yet most detection systems oversimplify the task by focusing on binary classification, ignoring nuanced categories like passive-aggressive remarks or indirect slurs. To address this gap, we propose a hybrid framework combining Term Frequency-Inverse Document Frequency (TF-IDF), word-to-vector (Word2Vec), and Bidirectional Encoder Representations from Transformers (BERT) based models for multi-class cyberbullying detection. Our approach integrates TF-IDF for lexical specificity and Word2Vec for semantic relationships, fused with BERT’s contextual embeddings to capture syntactic and semantic complexities. We evaluate the framework on a publicly available dataset of 47,000 annotated social media posts across five cyberbullying categories: age, ethnicity, gender, religion, and indirect aggression. Among BERT variants tested, BERT Base Un-Cased achieved the highest performance with 93% accuracy (standard deviation ±1% across 5-fold cross-validation) and an average AUC of 0.96, outperforming standalone TF-IDF (78%) and Word2Vec (82%) models. Notably, it achieved near-perfect AUC scores (0.99) for age and ethnicity-based bullying. A comparative analysis with state-of-the-art benchmarks, including Generative Pre-trained Transformer 2 (GPT-2) and Text-to-Text Transfer Transformer (T5) models highlights BERT’s superiority in handling ambiguous language. This work advances cyberbullying detection by demonstrating how hybrid feature extraction and transformer models improve multi-class classification, offering a scalable solution for moderating nuanced harmful content.

---

Keywords

---