<![CDATA[The advancement of digital technology has driven the increasing use of job search applications such as KitaLulus. User reviews on the Google Play Store serve as a crucial source for evaluating service quality and user satisfaction. This study aims to analyze user sentiment toward the KitaLulus application using the Support Vector Machine (SVM) algorithm, combined with the Synthetic Minority Over-sampling Technique (SMOTE) to address class imbalance in sentiment data. The research process includes collecting 1,000 user reviews through web scraping, text preprocessing, rating-based labeling, data transformation using TF-IDF, splitting the dataset into 80% training and 20% testing, applying SMOTE, training the SVM model, and evaluating its performance. The results show that SVM trained with SMOTE-balanced data achieved an accuracy of 89%, precision of 90%, recall of 89%, F1-score of 90%, and an AUC of 0.93. This study contributes a practical implementation of the SVM-SMOTE combination, demonstrating its effectiveness in text-based sentiment classification, particularly in handling imbalanced review data from mobile applications.]]>
Copyrights © 2025
