Analisis Sentimen Ulasan Aplikasi RedBus Menggunakan Metode SVM dan AdaBoost

Abstract

Bus transportation is a mode of transportation relied upon by the public due to its accessibility and affordable ticket prices. Sentiment analysis of RedBus app reviews on the Google Play Store can provide insight into user sentiment toward the app. The aims of this study are to analyze sentiment in reviews of the RedBus app using two approaches: the Random Forest model and a combination of Random Forest with AdaBoost. The analysis classifies user opinions as positive or negative. The study uses the TF-IDF method for feature extraction, and the evaluation methods include K-Fold Cross Validation and Confusion Matrix. The findings indicate that the combination of Random Forest with AdaBoost significantly enhances performance compared to the standard Random Forest model. Using a combination of Random Forest with AdaBoost results in an average accuracy of 89.0%, while the standard Random Forest model attains an average accuracy of 85.0%.

Keywords: Sentiment Analysis, RedBus, Random Forest, AdaBoost, TF-IDF