MACHINE LEARNING-BASED FAKE NEWS DETECTION FOR SOCIAL MEDIA
Keywords:
Fake News Detection, Machine Learning, Natural Language Processing, Uzbek Social Media, Information Security, Text ClassificationAbstract
With the rapid growth of social media platforms such as Telegram, Facebook, and Instagram, the spread of false, unverified, and manipulative information (fake news) in the Uzbek language has increased significantly, causing social anxiety, misinformation, and public misunderstanding. Existing international fake news detection algorithms are primarily designed for the English language, and due to the rich morphological structure, agglutinative nature of Uzbek, and the scarcity of labeled datasets, direct application of these algorithms yields inefficient results. This paper proposes a hybrid system for automatically determining the veracity of Uzbek-language texts and headlines using TF-IDF and a multilingual BERT (mBERT) model adapted for Uzbek. The system combines classical machine learning algorithms — Logistic Regression, Random Forest, and XGBoost — with transformer-based contextual embeddings. A dataset of more than 2,500 manually labeled Uzbek news articles and social media posts was collected, followed by preprocessing steps including tokenization, lemmatization, and stop-word removal. Expected results indicate that the mBERT-based model will demonstrate higher Accuracy, Precision, Recall, and F1-Score metrics compared to classical algorithms. The proposed system contributes to ensuring information security, combating disinformation, and enriching NLP resources for the Uzbek language.References
1. Sharma, A. et al. (2023). A Comprehensive Survey on Fake News Detection with Machine Learning Algorithms. IEEE Xplore. https://ieeexplore.ieee.org/document/10270611/
2. A Survey on Role of Machine Learning and NLP in Fake News Detection on Social Media. IEEE Conference Publication, 2021. https://ieeexplore.ieee.org/document/9573875/
3. Fake News Detection Using Machine Learning Algorithm (XGBoost-based approach). IEEE Conference Publication, 2023. https://ieeexplore.ieee.org/document/10088320/
4. Fake News Detection using Machine Learning (TF-IDF and SVM approach). IEEE Conference Publication, 2021. https://ieeexplore.ieee.org/document/9358890/
5. Devlin, J., Chang, M.-W., Lee, K., Toutanova, K. (2018). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv preprint arXiv:1810.04805. https://arxiv.org/abs/1810.04805
6. Multilingual Fake News Detection: A Study on Various Models and Training Scenarios (mBERT, XLM-RoBERTa, LASER). Springer Nature Link, 2024. https://link.springer.com/chapter/10.1007/978-3-031-66428-1_5
7. A Transformer-Based Approach to Multilingual Fake News Detection in Low-Resource Languages. ACM Transactions on Asian and Low-Resource Language Information Processing, 2021. https://dl.acm.org/doi/10.1145/3472619
8. Kuriyozov, E. et al. (2023). Text Classification Dataset and Analysis for Uzbek Language. arXiv preprint arXiv:2302.14494. https://arxiv.org/abs/2302.14494
9. Rabbimov, I., Kobilov, S. Multi-class Text Classification of Uzbek News Articles using Machine Learning, as cited in Kuriyozov, E. et al., Text Classification Dataset and Analysis for Uzbek Language, arXiv:2302.14494, 2023. https://arxiv.org/pdf/2302.14494
10. Mansurov, B., Mansurov, A. (2021). UzBERT: Pretraining a BERT Model for Uzbek. arXiv preprint arXiv:2108.09814. https://arxiv.org/abs/2108.09814
11. BERTbek: A Pretrained Language Model for Uzbek. Proceedings of SIGUL 2024 (ACL Anthology). https://aclanthology.org/2024.sigul-1.5.pdf
12. A Feedforward Neural Model for Morphological Stemming of Uzbek Words. ResearchGate, 2025.