SENTIMENT ANALYSIS OF TEXT USING BERT AND DISTILBERT MODELS

Abstract

The work describes the fine-tuning of BERT and DistilBERT pre-trained transformer models for the binary sentiment classification of user comments in the Ukrainian language. The models were evaluated according to accuracy, precision, recall, and F1-score metrics. An overall accuracy of 91% obtained for both models demonstrates the significant potential of incorporating these models into sentiment analysis tasks.