Digital Korlantas POLRI is an application released on April 13, 2021 by Korlantas POLRI. This application aims to provide SIM extension services to the public digitally. This application has reached more than 5 million downloads, with 102 thousand reviews and a rating of 3.7. The reviews include a variety of positive and negative coments. Many users complained about the app’s performance. Therefore, sentiment analysis is needed with the aim of utomatically classifying a person’s sentiments into positive and negative groups. This research uses the Naïve Bayes algorithm as a classification method with a dataset of 1500, consisting of 536 positive sentiment data and 964 negative sentiment data. In a data processing process using data division with a ratio of 80:20, resulting in 1200 training data and 300 test data. The final result of this research is to get accuracy, precision and recall values which aim to evaluate how well the Naïve Bayes algorithm performs in the classification process. The values obtained from the evaluation by using data as much as 300 training data, shows an accuracy value of 70.33%, a precision value of 56.34%, and a recall value of 74.77%. The evalution test results show that the Naïve Bayes algorithm is successful in the classification process.
Copyrights © 2024