Development of an Artificial Intelligence Model for Detecting Toxic (Harmful) Content in Texts
Received: 2026-05-30
Published: 2026-06-06
Abstract
This paper addresses the problem of automatic detection of toxic and harmful content in texts. The proliferation of toxic content on social networks and online platforms has a negative impact not only on users but also on society as a whole. The study proposes a hybrid artificial intelligence architecture that combines BERT-based transformers with bidirectional LSTM networks. The proposed model was trained on mixed Uzbek and Russian texts collected from Uzbek social media platforms. Experiments showed that the model achieved 94.7% accuracy, 93.2% recall, and 94.0% F1-score. In addition, the paper analyzes dataset preparation, model evaluation metrics, and real-time deployment capabilities.
Keywords
List of references
-
Nobata, C., Tetreault, J., Thomas, A., Mehdad, Y., & Chang, Y. (2016). Abusive Language Detection in Online User Content. Proceedings of WWW 2016, 145–153.
-
Davidson, T., Warmsley, D., Macy, M., & Weber, I. (2017). Automated Hate Speech Detection and the Problem of Offensive Language. Proceedings of ICWSM 2017, 512–515.
-
Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. Proceedings of NAACL-HLT 2019, 4171–4186.
-
Zhang, Y., Wallace, B. (2018). A Sensitivity Analysis of (and Practitioners' Guide to) Convolutional Neural Networks for Sentence Classification. Proceedings of IJCNLP, 253–263.
-
Zampieri, M., Malmasi, S., Nakov, P., Rosenthal, S. (2019). Predicting the Type and Target of Offensive Posts in Social Media. Proceedings of NAACL-HLT 2019, 1415–1420.
-
Kovačević, A., Dragović, I., Kosmajac, D. (2021). Offensive Language Detection in Serbian Texts Using Fine-Tuned BERT. Applied Sciences, 11(12), 5611.
-
Mulki, H., Zaghouani, W. (2021). HabiRemoved: Hate Speech and Offensive Language Detection for Arabic. ACL Anthology Workshop Proceedings.
-
Mamatov, A., Yusupov, K., Tursunov, O. (2020). Developing Word Embeddings for Uzbek Language Processing. Uzbek Journal of Information Technologies, 4(2), 31–44.
-
Yusupov, B., Khoroshevsky, V. (2022). Social Media Text Analysis for Uzbek Language: Challenges and Opportunities. Proc. TürkLang 2022, 88–97.
-
Shokh, A., Yulchiev, D. (2023). Code-Switching in Uzbek Digital Communication: Linguistic Patterns and Computational Challenges. Central Asian Computational Linguistics Conference, 14–22.
About the Authors
License
This work is licensed under a Creative Commons Attribution 4.0 International License.