<![CDATA[The increasing number of user reviews on the Google Play Store provides valuable information for app developers in evaluating service quality. However, the increasing number of reviews makes manual analysis less effective. This study aims to develop a near real-time sentiment analysis system for PineDrama app reviews using the Support Vector Machine (SVM) method with Term Frequency–Inverse Document Frequency (TF-IDF) weighting. The system is designed to automatically retrieve the latest reviews using Google Play Scraper, then perform a preprocessing process that includes case folding, cleaning, normalization, stopword removal, and stemming. Next, the data is classified into positive, neutral, and negative sentiments using a trained SVM model. The classification results are stored in a SQLite database and displayed through a Flask-based dashboard that is updated periodically using a scheduler. Based on the test results, the SVM model produces an accuracy of 91.06%, with a precision value of 92.50%, a recall of 91.06%, and an F1-score of 91.50%, indicating that the model has good performance in classifying review sentiments. The developed system also successfully displays analysis results in near real-time, thus helping the process of monitoring user opinions on the PineDrama application more quickly and efficiently.]]>
Copyrights © 2026
