<![CDATA[The rapid growth of mobile applications in Indonesia has generated large volumes of user reviews that can be used to evaluate service quality and user experience. However, Indonesian-language reviews are often informal, noisy, and context dependent, making sentiment classification a challenging Natural Language Processing task. This study aims to compare the performance of Random Forest and Support Vector Machine in classifying Indonesian mobile application reviews into positive, neutral, and negative sentiment classes. The proposed research pipeline consists of data collection, text cleaning, case folding, slang normalization, tokenization, stop-word removal, stemming, TF-IDF feature extraction, model training, and evaluation using accuracy, precision, recall, F1-score, and confusion matrix. In the example experimental scenario involving 3,000 balanced reviews and an 80:20 train-test split, Support Vector Machine obtained higher accuracy and macro F1-score than Random Forest, namely 89.33% and 0.887 compared with 86.50% and 0.861. The result indicates that SVM is more suitable for sparse high-dimensional TF-IDF features, while Random Forest remains useful as an interpretable ensemble baseline. This study contributes a reproducible comparative framework for Indonesian mobile-app sentiment classification and may support developers in transforming user reviews into data-driven improvement priorities.]]>
Copyrights © 2026
