<![CDATA[The rapid growth of generative artificial intelligence applications, particularly ChatGPT, has resulted in a significant increase in user reviews on the Google Play Store. These reviews serve as valuable sources for understanding user perceptions, experiences, and concerns. This study aims to analyze sentiment in Indonesian-language reviews of the ChatGPT application using the Bidirectional Encoder Representations from Transformers (BERT) algorithm combined with the Knowledge Discovery in Database (KDD) methodology. The dataset was collected using web scraping via the google-play-scraper library, producing 1,806 reviews after data cleaning and preprocessing. The dataset was divided into training and testing sets with an 80:20 ratio. IndoBERT was employed as the pre-trained model. Evaluation results show that the model successfully classified positive, negative, and neutral sentiments with an accuracy of 93%, precision of 87%, recall of 83%, and an F1-score of 85%. Although performance on the neutral class was lower due to dataset imbalance, the model demonstrated strong overall results. This study confirms that BERT is effective for sentiment analysis of application reviews and can serve as a reference for improving application service quality by understanding user opinions.]]>
Copyrights © 2025
