<![CDATA[In this modern era, the use of technology in various fields of life is growing, especially in the financial sector such as digital payments. Digital payments are transactions made via the internet with various forms or electronic payment applications. This research utilizes one of the e-wallet applications, namely the DANA Application. As one of the digital payment applications, DANA's service quality is an important factor in competing with various e-wallet applications or other competitive digital payment applications, sentiment analysis of massive amounts of incoming review data is needed to understand user perceptions and satisfaction levels in order to help companies in decision making. This research analyzes 69632 user review data from May 1 to June 28, 2024, then classifies reviews into positive, negative, and neutral reviews based on aspects to find out the dominant aspects in the review data, as for the aspects determined in this study are mobile applications, interfaces, service performance, and security. This research compares various classification method algorithms to determine the performance of the model, the comparison results show that the Random Forest algorithm has the best performance of other algorithms studied based on evaluation metrics with accuracy 0.82, precision 0.82, recall 0.82, f1_score 0.82, AUC-ROC 0.92, and Cross Validation Mean Accuracy 0.822. This research is expected to produce useful and computerized knowledge extraction and provide perceptual understanding, user satisfaction, and insight into the most effective algorithms which can then help companies in making decisions in the future.]]>
Copyrights © 2024
