Sentiment Analysis of KA Bandara Application Using Support Vector Machine and Random Forest Methods

Analisis Sentimen Aplikasi KA Bandara Mengunakan Metode Support Vector Machine dan Random Forest

  • Muhammad Hilmy Maulana no
Keywords: Sentiment Analysist, TF-IDF, Support Vector Machine, Random Forest, Web Scraping

Abstract

 Digital transformation has significantly impacted public transportation services, including the adoption of the KA Bandara application that facilitates ticket booking and travel information access. User reviews on the Google Play Store serve as a valuable source for analyzing user perceptions regarding service quality. This study performs sentiment analysis on 1,000 KA Bandara app reviews using two classification algorithms: Support Vector Machine (SVM) and Random Forest (RF). The review data were collected via web scraping and preprocessed through steps such as case folding, tokenization, stopword removal, and stemming. Text features were extracted using TF-IDF, and sentiment labels were assigned based on star ratings (positive: 4–5; negative: 1–3). The models were evaluated using accuracy, precision, recall, and F1-score metrics. The results indicate that SVM outperforms RF, achieving an accuracy of 89%, while RF recorded 86.5%. This research confirms the superior performance of SVM for sentiment classification of digital app reviews. The findings offer actionable insights for developers to enhance service quality by leveraging user feedback. 

Downloads

Download data is not yet available.
Published
2025-07-24
How to Cite
Maulana, M. (2025, July 24). Sentiment Analysis of KA Bandara Application Using Support Vector Machine and Random Forest Methods. PROSIDING SEMINAR NASIONAL SAINS DATA, 5(1), 370-379. https://doi.org/https://doi.org/10.33005/senada.v5i1.491