<![CDATA[This research aims to evaluate user responses to the Indodana: Paylater & Pinjaman application through sentiment analysis using the naive bayes algorithm. Online lending apps such as Indodana have changed the way individuals access finance by providing a quick and easy process. However, the user's decision to choose a legal app and pay attention to the transparency of fees and loan terms is crucial. With more than ten million downloads and two million reviews, it is important to understand user sentiment so that developers can improve services and maintain public trust. A sentiment analysis method using multinominal naive bayes was used with two labelling approaches inset lexicon and rating. The evaluation was conducted on 500 Indodana: Paylater & Pinjaman reviews, dividing the data into training and testing and using TF-IDF features. The results show that inset lexicon labelling achieved 86% accuracy, whereas rating-based labelling achieved 87% accuracy. These results provide an in-depth view of user responses, aiding in the identification of factors that influence positive or negative perceptions of the app. As such, this research is important for guiding the development of safe, reliable, and compliant online lending applications, as well as for improving overall user satisfaction]]>
Copyrights © 2024
