<![CDATA[Sentiment analysis on reviews of an application becomes an option to see users responses to the service of that particular application. Random Forest is one of the classification modeling techniques that originate from a combination of Decision Trees, providing the final result based on majority voting. This research aims to improve the performance of sentiment classification on customer reviews of Seabank, one of the most widely used digital banking services in Indonesia, by utilizing the Random Forest algorithm. The study involves sentiment analysis of user reviews on the Seabank application, collected from 15,000 reviews on Google Playstore. The review features available on Google Playstore are used as a means to convey opinions as user feedback for an application. Random Forest is trained to classify reviews into 3 sentiment classes: positive, neutral, and negative. Based on the research conducted with model evaluation using Confusion Matrix, an accuracy value of 94.1% was obtained, indicating that Random Forest's accuracy in classifying Seabank customer reviews is 94.1%. This demonstrates the effectiveness of using Random Forest in text review classification due to its high accuracy value.]]>
Copyrights © 2026
