A Hybrid Multi-Modal Model for Detecting Cyberbullying Online

DOI

10.2991/978-94-6239-610-4_43How to use a DOI?

Keywords

Cyberbullying Detection; Multi-Modal Learning; Deep Learning; CNN; RNN; Transformer Models; Text Analysis; Image-Based Harassment; Sentiment Analysis; User Behavior Modeling; Social Media Safety; Machine Learning; Online Abuse Identification; Real-Time Monitoring; Early Warning System

Abstract

Cyberbullying has emerged as a critical challenge in modern digital communication, severely impacting the psychological well-being of social media users, especially adolescents. Traditional detection techniques relying solely on textual features often fail to capture the complex, multi-dimensional nature of online harassment, which frequently includes images, emojis, slang, sentiment cues, and user behavioral patterns. To address these limitations, this research proposes a comprehensive multi-modal approach for the identification and detection of cyberbullying across social networking platforms. The proposed system integrates textual analysis, image interpretation, sentiment extraction, metadata patterns, and user interaction behavior to construct a robust and context-aware detection framework. Deep learning architectures such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Transformer-based models, and multi-modal fusion layers are utilized to learn complementary features from heterogeneous data sources. The system processes offensive language, harmful visual content, aggressive conversation patterns, and user profiling factors to improve detection accuracy. Experimental evaluation on benchmark datasets demonstrates that multi-modal learning significantly outperforms single-modal approaches by capturing hidden, implicit, and context-dependent bullying signals. Furthermore, the model supports early warning mechanisms and real-time monitoring to aid social media platforms in mitigating harmful interactions. This study highlights the importance of integrating diverse data modalities and advanced neural architectures to build reliable, scalable, and intelligent cyberbullying detection solutions, contributing to safer and more responsible online communities.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Patlola Sridhar
AU  - Meeravali Shaik
PY  - 2026
DA  - 2026/05/05
TI  - A Hybrid Multi-Modal Model for Detecting Cyberbullying Online
BT  - Proceedings of the First International Conference on Advances in Forensics and Cyber Technologies (ICFACT 2025)
PB  - Atlantis Press
SP  - 498
EP  - 506
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6239-610-4_43
DO  - 10.2991/978-94-6239-610-4_43
ID  - Sridhar2026
ER  -