Sentiment Analysis of Transjakarta User Feedback Using Machine Learning Models

The transjakarta bus rapid transit system plays a crucial role in Jakarta’s transportation network, serving millions of passengers annually. However, user feedback often highlights concerns regarding service quality, punctuality, and fleet cleanliness. To address these issues, this study conducts a sentiment analysis of user reviews from the Transjakarta app on the Google Play Store, aiming to evaluate public perception and provide actionable insights for service improvement. The research employs machine learning techniques, including Naïve Bayes, Support Vector Machine (SVM), and Decision Tree algorithms, to classify user sentiments into positive, negative, and neutral categories. The results indicate that the SVM model outperforms other algorithms in terms of classification accuracy, achieving a significant level of precision in identifying both positive and negative sentiments. These findings suggest that while most user feedback highlights dissatisfaction with punctuality and cleanliness, there are notable positive sentiments regarding the application’s usability. The study concludes that sentiment analysis is a powerful tool for understanding user opinions and can guide decision-making for enhancing service quality in public transportation systems. The insights gained from this research can be utilized by Transjakarta management to address service deficiencies, optimize customer satisfaction, and improve operational efficiency, thereby fostering a more effective public transportation system in Jakarta.