DANA User Review Sentiment Analysis with Machine Learning Algorithms

In the current digital age, digital financial platforms like DANA have gained significant popularity, playing a crucial role in simplifying transactions and financial management. User feedback on platforms such as Google Playstore provides valuable insights into user satisfaction and service perception. This study focuses on performing sentiment analysis on DANA app reviews using three common classification algorithms: Naive Bayes, Support Vector Machine (SVM), and Random Forest. The research process includes gathering review data from Google Playstore, labeling, data preprocessing, and applying sentiment analysis with the Naive Bayes, SVM, and Random Forest algorithms. The findings reveal that out of 4,329 user reviews, the majority showed neutral sentiment (1,429 reviews), followed by positive sentiment (1,309 reviews) and negative sentiment (1,328 reviews). The Random Forest algorithm delivered the highest accuracy at 94%, with SVM achieving 93%, and Naive Bayes 76%. In terms of computation time, Random Forest exhibited strong performance, completing in 29.25 seconds (training time: 29 seconds, prediction time: 0.25 seconds).