Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{187804,
        author = {Yeshodha R},
        title = {A Comprehensive Review of Cyberbullying Detection Techniques and Datasets on Social Media Platforms},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {6},
        pages = {6943-6959},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=187804},
        abstract = {Cyberbullying has emerged as a critical issue in today's digital society, particularly among adolescents who extensively use social media for communication and information exchange. Unfortunately, some users exploit these online platforms to harass, humiliate, or intimidate others through posts, messages, or digital interactions, resulting in severe psychological and emotional consequences for victims. Although numerous studies have investigated the detection, prevention, and mitigation of cyberbullying, the problem persists due to evolving online behaviors and linguistic complexities. This study presents a comprehensive systematic review of existing research on cyberbullying detection from 2018 to 2024. It examines state-of-the-art datasets, methodologies, and technologies—ranging from traditional machine learning models to deep learning and large language models (LLMs). The paper identifies key challenges, research gaps, and opportunities for future exploration, offering insights and recommendations to enhance the accuracy, fairness, and scalability of cyberbullying detection systems across diverse social media platforms. The review analyzed 27 papers using the PRISMA framework, revealing that hybrid deep learning models (CNN-BiLSTM) achieve the highest performance (93.4% accuracy, 92.8% F1-score), while dataset imbalance and English-language dominance (74.1%) remain critical challenges.},
        keywords = {Cyberbullying Detection, Social Media Analysis, Machine Learning, Deep Learning, Natural Language Processing, Large Language Models, Online Harassment, Hate Speech Detection, Sentiment Analysis, Dataset Imbalance.},
        month = {November},
        }