This study aims to identify the sentiment of user reviews of the POLRI Super App on Google Play Store using the Multinomial Naive Bayes algorithm. The application is a digital service provided by the Indonesian National Police; however, its effectiveness and user satisfaction levels still require further examination. The data used in this study consist of user reviews obtained directly from the Google Play platform. The analysis process began with text preprocessing stages such as tokenization, stopword removal, and stemming, followed by sentiment labeling into three categories: positive, negative, and neutral. The Naive Bayes model was trained using an 80% training set and a 20% testing set. Model evaluation results show that this algorithm achieved an accuracy rate of 87%, with the best performance in classifying neutral sentiments. These findings indicate that the Multinomial Naive Bayes is quite effective for sentiment classification in short review texts. This research is expected to serve as a reference for developing public opinion monitoring systems based on user reviews of digital public services.
Candra Zaenudin Zidane (Wed,) studied this question.