International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | Cyberbullying Detection on Social Platform Using TF-IDF and SVM |
|---|---|
| ABSTRACT | Cyberbullying has become a serious concern in modern digital communication, particularly across social media platforms, online forums, gaming communities, and messaging applications. The increasing volume of user-generated content and the use of images, memes, and voice messages make manual moderation impractical and traditional text-only systems inadequate. This paper proposes a multimodal cyberbullying detection system that examines text, image, and audio inputs to detect abusive and harmful content. Textual information is handled using TF-IDF and BERT embeddings, and text contained in images is extracted using Optical Character Recognition (OCR) techniques. Audio inputs are translated into text using speech-to-text methods prior to classification. Support Vector Machine and Logistic Regression classifiers are employed for prediction. Besides binary classification, the system proposes the Emotional Damage Score (EDS) to approximate the level of harmful content, providing improved interpretability for moderation and intervention. A web interface built with Flask enables users to upload content and display outputs, and an administrative interface facilitates monitoring, feedback, error correction, and controlled model updating. Experimental findings show that the proposed method enhances contextual understanding and detection performance compared to traditional text-only systems. The system demonstrates practical applicability for deployment in educational institutions and online platforms aimed at promoting safer digital interactions. |
| AUTHOR | KUSUMA C, MOHAN P, N WILSON PREM KUMAR, NAVEEN N C, RAKSHITH RAJATH B R Assistant Professor, Department of Information Science and Engineering, S J B Institute of Technology, Bengaluru, India Department of Information Science and Engineering, S J B Institute of Technology, Bengaluru, India |
| VOLUME | 184 |
| DOI | DOI: 10.15680/IJIRCCE.2026.1405058 |
| pdf/58_Cyberbullying Detection on Social Platform Using TF-IDF and SVM.pdf | |
| KEYWORDS | |
| References | [1] S. Sharma, P. Yadav, and R. Kaur, “A transformer-based approach for cyberbullying detection on social media,” IEEE Access, vol. 12, pp. 15432-15445, 2024. [2] A. Banerjee, M. Khan, and R. Patel, “Context-aware bullying classification using BERT embeddings,” Int. J. Adv. Comput. Sci. Appl., vol. 15, no. 10, pp. 91-99, 2024. [3] M. Garg and S. Gupta, “Text mining and TF-IDF based offensive language detection in online conversations,” Procedia Comput. Sci., vol. 232, pp. 145–154, 2023. [4] R. Singh and V. Chauhan, “Multimodal hate speech detection in memes using OCR and deep neural networks,” Appl. Intell., vol. 54, no. 8, pp. 9123-9140, 2024. [5] J. Li, T. Zhang, and A. Roy, “Enhanced cyberbullying prevention using SVM and logistic regression classifiers,” Comput. Electr. Eng., vol. 110, pp. 108-122, 2024. [6] K. Deshmukh, F. Rahman, and J. Silva, “Speech-to-Text abusive speech recognition for online gaming safety,” IEEE Trans. Affect. Comput., vol. 14, no. 3, pp. 545-558, 2025. [7] L. Wei and H. Zhou, “Optical character recognition-based offensive content detection in social multimedia,” Pattern Recognit. Lett., vol. 176, pp. 16-24, 2024. [8] A. Rao and N. Bansal, “An integrated NLP-based framework for cyberbullying analysis using deep learning,” Expert Syst. Appl., vol. 241, art. 122673, 2024. [9] M. S. Ali, S. Verma, and R. Mehta, “Measuring psychological impact of harassment using supervised learning and severity scaling,” J. Inf. Secur. Appl., vol. 78, art. 103653, 2025. [10] P. Chauhan, V. Joshi, and S. Taneja, “A systematic review of machine learning approaches for cyberbullying detection,” IEEE Rev. Comput. Softw. Eng., vol. 29, pp. 411-435, 2024. [11] Google AI Research, “BERT: Pre-training of deep bidirectional transformers for language understanding,” Proc. NAACL-HLT, pp. 4171-4186, 2019. [12] R. Smith, “An overview of the Tesseract OCR engine,” Proc. Int. Conf. Document Anal. Recognit., pp. 629–633, 2018. [13] J. Kim, “Automatic Speech Recognition based on hybrid deep learning systems,” IEEE Signal Process. Mag., vol. 40, no. 5, pp. 55-70, 2023. [14] M. Miller and D. Lee, “Building secure machine learning Web APIs using Flask and SQLAlchemy,” Software Pract. Exper., vol. 55, no. 1, pp. 88-102, 2024. [15] A. Singh and P. S. Rao, “Web-based real-time abusive speech moderation system using multimodal analysis,” Front. Artif. Intell., vol. 7, art. 1432857, 2024. [16] R. Verma, D. Srivastava, and P. Kulkarni, “Hybrid CNN–BiLSTM architecture for cyberbullying detection in multilingual social networks,” Neural Comput. Appl., vol. 36, no. 2, pp. 2214–2230, 2025. [17] S. Ahmed and G. Thomas, “Deep contextual sentiment modeling for aggressive behavior classification in online forums,” J. Big Data Analytics, vol. 12, no. 4, pp. 304–320, 2024. [18] N. Fernandes, K. Prakash, and Y. Li, “Graph neural networks for relational bullying behavior detection across social platforms,” Expert Syst., vol. 41, no. 5, art. e13562, 2025. [19] P. Reddy and M. Das, “Transformers with contrastive learning for abusive language detection on short text messaging platforms,” Comput. Intell., vol. 40, no. 1, pp. 98–115, 2024. [20] H. Qureshi, L. Wang, and S. Banik, “Real-time streaming cyberbullying alert system using cloud-based microservice architecture,” IEEE Internet Comput., vol. 29, no. 3, pp. 55–67, 2025. |
