Sentiment Analysis on User Reviews of the Edlink Application Using the Random Forest Classifier Method

Edlink is a learning platform developed by PT. Sentra Vidya Utama (SEVIMA), established in 2004. Although it offers useful features, some aspects need improvement based on user reviews on Google Play Store. This study aims to accurately classify user sentiment to identify areas that need enhancement. The main challenges include language diversity, sentiment class imbalance, and the need for a reliable classification method. The random forest classifier method was chosen for its ability to handle overfitting and optimize performance. The dataset consists of 1,117 reviews divided into three classes: 385 negative, 118 neutral, and 614 positive. Data was collected through web scraping and processed using cleaning, normalization, tokenizing, stemming, negation conversion, and stopword removal, then weighted using TF-IDF. Testing results showed an accuracy of 86% using 5-Fold cross-validation and SMOTE. The 10-Fold cross-validation test demonstrated that this method outperforms other classification methods with 90% accuracy.