ANALISIS SENTIMEN PUBLIK TERHADAP APLIKASI MYPERTAMINA PADA GOOGLE PLAYSTORE

Abstract

This study aims to analyze public sentiment towards the MyPertamina application on Google Playstore in order to understand user perceptions and identify the main problems faced. The method used is machine learning-based sentiment analysis with the optimized Support Vector Machine (SVM) algorithm. Data was obtained from user reviews on Google Playstore within a certain time period, then pre-processed and classified sentiment into positive, negative, or neutral. The results showed that the majority of reviews were positive, but there were a number of negative reviews that reflected technical and policy challenges in using the application. The SVM model managed to achieve an accuracy of 85.31%, indicating effectiveness in identifying public sentiment. These results are expected to help PT Pertamina in improving the quality of MyPertamina application services.