A Multimodal Cyberbullying Detection Framework for E-Commerce Platforms Using Hybrid Optimization Models
Keywords:
Cyberbullying Detection, E-Commerce Platforms, Multimodal Learning, Hybrid Optimization Models, Deep Learning, Online Harassment, Sentiment Analysis.Abstract
The fast development of e-commerce sites has highly changed the nature of online consumer relations where users are able to share reviews, comments, messages and multimedia contents in real-time. Although it improves the engagement of the users, this has given way to an alarming increase in negative behaviors attributable to cyberbullying like hate speech, negative reviews, targeted harassment, and toxic multimedia content. Some of the classic methods of cyberbullying detection eventually use the simplistic unimodal analysis of text and may not be able to detect the multimodal aspects of Web-based mistreatment found in contemporary e-commerce environments. In order to overcome these problems, the paper will suggest a multimodal cyber bullying detection system, which analyzes text, visual and contextual data with the help of hybrid optimization models. The framework proposed joins the deep learning approach with the features extraction and metaheuristic optimization methods to enhance the accuracy, robustness and scalability of detection. Transformer-based language models are used to extract their textual features, convolutional neural networks to capture their visual features through images, and additional features like user behavior and a trail of user interaction are also used to improve the semantic meaning. Hybrid optimization algorithms are to fine-tune model parameters as well as optimize feature fusion strategies. It is proven through experimental assessment on benchmark and real-world e-commerce data that the proposed framework is much more efficient than the traditional methods of the unimodal and non-optimized multimodal in terms of accuracy, precision, recall and F1-score. Although the performance is promising, the framework has practical constraints in the form of data privacy, cost of computation and real time implementation. The research directions in the future will be the lightweight model design, cross-linguistic cyberbullying detection, explainable artificial intelligence integration, and privacy-preserving learning method to facilitate ethical and scalable use in real-life e-commerce applications.
Downloads
Downloads
Published
Issue
Section
License
You are free to:
- Share — copy and redistribute the material in any medium or format for any purpose, even commercially.
- Adapt — remix, transform, and build upon the material for any purpose, even commercially.
- The licensor cannot revoke these freedoms as long as you follow the license terms.
Under the following terms:
- Attribution — You must give appropriate credit , provide a link to the license, and indicate if changes were made . You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
- No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.
Notices:
You do not have to comply with the license for elements of the material in the public domain or where your use is permitted by an applicable exception or limitation .
No warranties are given. The license may not give you all of the permissions necessary for your intended use. For example, other rights such as publicity, privacy, or moral rights may limit how you use the material.

