Real-Time Browser-Based Early Warning System for Cyberbullying Detection in Onlines

Notice

This is an unedited manuscript accepted for publication and provided as an Article in Press for early access at the author’s request. The article will undergo copyediting, typesetting, and galley proof review before final publication. Please be aware that errors may be identified during production that could affect the content. All legal disclaimers of the journal apply.

Year : 2026 | Volume : 04 | 01 | Page :
    By

    Sonal Sharma,

  • Manish Singh,

  • Pratham Singh,

  1. Associate Professor, Department of MCA, Thakur Institute of Management Studies, Career Development & Research (TIMSCDR) Mumbai, Maharashtra, India
  2. Research Scholar, Department of MCA, Thakur Institute of Management Studies, Career Development & Research (TIMSCDR) Mumbai, Maharashtra, India
  3. Research Scholar, Department of MCA, Thakur Institute of Management Studies, Career Development & Research (TIMSCDR) Mumbai, Maharashtra, India

Abstract

The rise in social networking through Internet- based communication tools, Instagram, and YouTube, to name a few, significantly increases the risk of cyberbullying, thereby increasing psychological trauma on users, especially children, through adverse emotional states like “anxiety, depression, etc.” For a long time, researchers have been enhancing detection tools to counter cyberbullying, but their ability to detect only after the fact, along with limited support for English-based architecture, is a major concern that also compromises user “privacy.” This study seeks to develop a browser-based real-time warning system that can identify potentially harmful or offending messages before they are transmitted. This study targets the development of a scheme that is multilingual, privacy-concerned, and enables users with warnings before posting messages that protect them from harm. The proposed methodology involves the usage of secondary data sets containing previously annotated data from social media sources like Instagram and YouTube, along with the survey method of questionnaire formation. Additionally, the proposed approach utilizes the lightweight multilingual transformer architecture for the detection of cyberbullying; the same architecture will be adapted for in-browser usage with the aid of TensorFlow.js so that the data never gets processed remotely from the client’s browser. The initial results also revealed the following insights regarding the competitive accuracy of the model, which could enable the reduction of inference latency: User responses also clearly pointed out the need for the model to cover the multilingual community, as well as the value of local data privacy. This study bridges the gap by using a combination of real, time detection, multi, linguistic support, and a browser, based solution for privacy, safe deployment. The suggested algorithm aims at making the internet environment safer by offering a better guard against cyber bullying risks on various social media platforms.

Keywords: Real time Detection, Browser based System, Privacy preserving AI, Natural Language Processing, Multilingual Analysis.

How to cite this article:
Sonal Sharma, Manish Singh, Pratham Singh. Real-Time Browser-Based Early Warning System for Cyberbullying Detection in Onlines. International Journal of Computer Science Languages. 2026; 04(01):-.
How to cite this URL:
Sonal Sharma, Manish Singh, Pratham Singh. Real-Time Browser-Based Early Warning System for Cyberbullying Detection in Onlines. International Journal of Computer Science Languages. 2026; 04(01):-. Available from: https://journals.stmjournals.com/ijcsl/article=2026/view=247708


References

  1. Hosseinmardi S, Mattson SA, Rafiq RI, Han R, Lv Q, Mishra S. Detection of cyberbullying incidents on Instagram. In: Proc Int Conf Social Informatics; 2015.
  2. Hosseinmardi S, Ghasemianlangroodi A, Han R, Lv Q, Mishra S. Prediction of cyberbullying on Instagram. In: Proc ACM Web Sci Conf (WebSci); 2015. p. 85–94.
  3. Liu P, Guberman J, Hemphill L, Culotta A. Forecasting hostility in Instagram comments. In: Proc Int AAAI Conf Web Soc Media (ICWSM); 2018. p. 100–9.
  4. Balakrishnan V, Ng K. Personality and emotions in cyberbullying detection on YouTube. Telemat Inform. 2018;35(5):1366–76.
  5. Ashraf A, Rizwan S, Imran M. Abusive language detection on YouTube with context. In: Proc IEEE Int Conf Adv Comput Commun Inform (ICACCI); 2021.
  6. Chatzakou K, Kourtellis N, Leontiadis I. Detection and prevention of cyberbullying in social media. In: Companion Proc Web Conf (WWW); 2021.
  7. Zhang Y, Li M, Xu H. Real-time cyberbullying detection using NLP. Expert Syst Appl. 2022;192:116124.
  8. Raj R, Kumar S, Srinivasan P. Cyberbullying detection in Indian languages. In: Proc Forum Inf Retrieval Eval (FIRE); 2022.
  9. Saha S, Mandal A. Multilingual offensive language detection using transformers. In: Proc Int Conf Lang Resour Eval (LREC); 2023. p. 455–63.
  10. Fitro A, Pradana M, Nurhadi R. Instagram cyberbullying detection with IndoBERTa. In: Proc ACL Asia; 2023.
  11. Kowsari K, Brown D, Heidarysafa J. Automatic detection of cyberbullying in text. IEEE Trans Affect Comput. 2021;12(3).
  12. Gupta R, Kumar V. AI-powered content moderation in social media. In: Proc Int Joint Conf Artif Intell (IJCAI); 2021.
  13. Kim J, Park H, Lee D. Privacy-preserving cyberbullying detection. In: Proc AAAI Conf Artif Intell; 2024.
  14. Banerjee S, Singh R. Explainable AI for cyberbullying detection. In: Proc Int Conf Comput Linguist (COLING); 2022.
  15. Harshitha N, Varma R, Rao T. ProTect: a hybrid cyberbullying detection model for Instagram and Twitter. In: Proc Int Conf Mach Learn (ICML); 2024.
  16. Ahmad A, Sharma P, Khan Y. Cross-platform cyberbullying detection framework. In: Proc ACM Web Sci Conf (WebSci); 2023.
Ahead of Print Subscription Review Article
Volume 04
01
Received 27/03/2026
Accepted 31/03/2026
Published 05/04/2026
Publication Time 9 Days
2

Login


My IP

PlumX Metrics