Enhancing User Satisfaction on a Digital Payment Platform Using Topic Modeling and Machine Learning Techniques

This study aims to predict user satisfaction in the GoPay application by integrating Topic Modeling and machine learning techniques. The research focuses on analyzing customer reviews and ratings to identify key themes influencing satisfaction levels. Latent Dirichlet Allocation (LDA) extracts topics such as user experience, app performance, feature usage, and customer support. These topics are then used as input for machine learning classifiers, including Random Forest, Support Vector Machine (SVM), Naive Bayes, and XGBoost, to predict satisfaction outcomes. Among these, Support Vector Machine (SVM), achieved the highest accuracy of 91,24%, demonstrating its reliability for this task. The findings provide valuable insights into customer sentiments and suggest practical improvements for GoPay’s services. This research also highlights the potential of machine learning methods to enhance user experiences in digital payment platforms and offers a foundation for further studies in fintech and recommendation systems.