<![CDATA[The rising adoption of digital wallets in Indonesia has created an urgent need for fintech service providers to gain deeper insights into user perceptions and satisfaction. However, many providers continue to rely on conventional methods for evaluating consumer feedback, without leveraging the power of machine learning–based sentiment analysis. This study aims to identify user sentiment patterns toward five leading digital wallet platforms GoPay, OVO, DANA, LinkAja, and ShopeePay and to evaluate the effectiveness of the Naïve Bayes algorithm in classifying sentiments based on user reviews from the Google Play Store. A quantitative approach was employed, involving web scraping, text preprocessing, and sentiment classification using the Multinomial Naïve Bayes algorithm. The results indicate that Naïve Bayes can achieve classification accuracy exceeding 85% across all platforms, with the highest performance observed for LinkAja (95.10%) and GoPay (93.33%). These findings reveal variations in public perception across platforms and underscore the potential of machine learning in supporting user voice–based digital service strategies. The novelty of this research lies in the integration of organic review data with predictive approaches to inform strategic decision-making in the fintech service context.]]>
Copyrights © 2026
