Using the RoBERTa and T5 Models for Analyzing Texts in Uzbek: NER Tasks and Punctuation Prediction

In recent years, Large Language Models have become increasingly popular due to their ability to solve a wide range of Natural Language Processing tasks. These models, such as RoBERTa and T5, demonstrate outstanding results in Named Entity Recognition and punctuation prediction tasks, which makes them especially useful for social media users. In this paper, we investigate the application of the RoBERTa and T5 models for NER tasks and punctuation prediction in the Uzbek language. We conduct a comparative analysis of the performance of these models, evaluating their accuracy and effectiveness. The results of the study show that the RoBERTa and T5 models are able to significantly improve automatic text processing, improving the quality and convenience of user interaction with social networks. This contribution is important for the further development of Natural Language Processing tools for Uzbek and other languages with limited resources.