The Application of BERT in Sentiment Analysis of IMDB Movie Reviews

This study aims to conduct a sentiment analysis of user reviews of the IMDb website using a fine-tuned BERT model. This approach uses review data, pre-processing data, fine-tuning of the BERT model, and model performance evaluation. This sentiment analysis uses secondary data taken from the Kaggle website to account for variations in public opinion on film reviews. The discussion of sentiment analysis findings revealed people's preferences in the form of positive sentiment in the storyline aspect, while negative sentiment revealed the duration aspect. The results showed that the BERT model achieved a high level of performance with an accuracy of 90%, precision of 89%, recall of 91%, and an F1-score of 90% on the validation dataset. The results of this test can be used by filmmakers to correct aspects that are not satisfactory to the audience in the next film production From the test results above, the BERT method can be used to conduct sentiment analysis with high accuracy, precision, recall, and f1-score test results.