DEVELOPMENT OF NEW BERT ALGORITHMS FOR NATIONAL TRANSLATION PROGRAMS BASED ON ARTIFICIAL INTELLIGENCE
Keywords:
Key words: BERT, machine translation, NLP, transformer, low-resource languages, Uzbek language, transfer learning, artificial intelligence, subword tokenization, cross-lingual learning.Abstract
Abstract. This paper explores the development and adaptation of advanced BERT-based algorithms for national machine translation systems. While transformer-based models have significantly improved the quality of neural machine translation, low-resource languages such as Uzbek still face substantial challenges due to limited parallel corpora, rich morphological structures, and domain diversity. The study analyzes recent improvements in BERT architectures, including multilingual pretraining, domain-specific fine-tuning, subword tokenization, and cross-lingual transfer learning. The paper proposes that optimized BERT-based frameworks can significantly enhance translation accuracy, contextual understanding, and semantic consistency in national translation programs. Experimental insights from recent NLP research indicate that hybrid architectures combining encoder-based BERT models with decoder-based transformer systems provide superior performance in low-resource settings.
Downloads
References
Devlin, J., Chang, M.W., Lee, K., Toutanova, K. (2018). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding.
Vaswani, A. et al. (2017). Attention Is All You Need. NeurIPS.
Conneau, A. et al. (2019). Unsupervised Cross-lingual Representation Learning at Scale (XLM-R).
Sutskever, I. et al. (2014). Sequence to Sequence Learning with Neural Networks.
Koehn, P. (2010). Statistical Machine Translation. Cambridge University Press.
Sennrich, R., Haddow, B., Birch, A. (2016). Neural Machine Translation of Rare Words with Subword Units.
Goodfellow, I., Bengio, Y., Courville, A. (2016). Deep Learning. MIT Press.
Sutton, R.S., Barto, A.G. (2018). Reinforcement Learning: An Introduction. MIT Press.
Johnson, M. et al. (2017). Google’s Multilingual Neural Machine Translation System.
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution 4.0 International License.
All content published in the Journal of Applied Science and Social Science (JASSS) is protected by copyright. Authors retain the copyright to their work, and grant JASSS the right to publish the work under a Creative Commons Attribution License (CC BY). This license allows others to distribute, remix, adapt, and build upon the work, even commercially, as long as they credit the author(s) for the original creation.
