DEVELOPING AN ARTIFICIAL INTELLIGENCE MODEL FOR DETECTING TOXIC (HARMFUL) CONTENT IN TEXTS

Authors

  • Toshpulatov Javoxir Xamidulla o‘g‘li Tashkent International University of Education Master’s student in Data Science National Committee of the Republic of Uzbekistan for Ecology and Climate Change Chief Specialist of the IT Department, “ECO-MEDIA CENTER” Author

Keywords:

Keywords: toxic content, artificial intelligence, natural language processing, machine learning, transformer, BERT, XLM-RoBERTa, low-resource languages, Uzbek language, content moderation.

Abstract

The article analyzes the theoretical and methodological foundations of building an artificial intelligence model for the automatic detection of toxic content in texts, namely insulting, degrading, threatening and hate-inciting statements. Taking into account the features of Uzbek as a low-resource language, the parallel use of the Latin and Cyrillic scripts and the phenomenon of code-switching with Russian, a conceptual model architecture for Uzbek-language texts is proposed.

References

1.O‘zbekiston Respublikasi Prezidentining 2021-yil 17-fevraldagi «Sun’iy intellekt texnologiyalarini jadal joriy etish uchun shart-sharoitlar yaratish chora-tadbirlari to‘g‘risida»gi PQ-4996-son qarori // Qonunchilik ma’lumotlari milliy bazasi. URL: https://lex.uz (accessed: 20.10.2026).

2.Davidson T., Warmsley D., Macy M., Weber I. Automated hate speech detection and the problem of offensive language // Proceedings of the International AAAI Conference on Web and Social Media. 2017. Vol. 11, No. 1. P. 512-515.

3.Sap M., Card D., Gabriel S., Choi Y., Smith N.A. The risk of racial bias in hate speech detection // Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics. Florence, 2019. P. 1668-1678.

4.Николенко С.И., Кадурин А.А., Архангельская Е.О. Глубокое обучение. Погружение в мир нейронных сетей. СПб.: Питер, 2018. 480 с.

5.Devlin J., Chang M.-W., Lee K., Toutanova K. BERT: pre-training of deep bidirectional transformers for language understanding // Proceedings of NAACL-HLT 2019. Minneapolis, 2019. P. 4171-4186.

6.Conneau A., Khandelwal K., Goyal N. et al. Unsupervised cross-lingual representation learning at scale // Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 2020. P. 8440-8451.

7.Smetanin S. Toxic comments detection in Russian // Computational Linguistics and Intellectual Technologies: Proceedings of the International Conference «Dialogue 2020». Moscow, 2020. Vol. 19. P. 1149-1159.

8.Dementieva D., Moskovskiy D., Logacheva V. et al. Methods for detoxification of texts for the Russian language // Multimodal Technologies and Interaction. 2021. Vol. 5, No. 9. Art. 54.

9.Kuriyozov E., Matlatipov S., Alonso M.A., Gómez-Rodríguez C. Construction and evaluation of sentiment datasets for low-resource languages: the case of Uzbek // Human Language Technology. Challenges for Computer Science and Linguistics. LTC 2019. Lecture Notes in Computer Science. Vol. 13212. Cham: Springer, 2022. P. 232-243.

10.Mansurov B., Mansurov A. UzBERT: pretraining a BERT model for Uzbek // arXiv preprint arXiv:2108.09814. 2021.

Downloads

Published

2026-11-02