<![CDATA[The JMO application is a key digital service platform developed by the Social Security Administration for Employment. This public legal entity provides social security protection for all workers in Indonesia, offering them online access. It offers a comprehensive suite of services, including participant registration, Old Age Security benefit simulations, balance checks, same-day claims with Electronic Know Your Customer (e-KYC) verification, and workplace accident reporting. This study aims to evaluate the application's usability by constructing an Indonesian-language corpus from user reviews on the Google Play Store, utilizing the Usability Metric for User Experience-Lite (UMUX-Lite) instrument. The objective is to assess whether the application meets user requirements and delivers a user-friendly experience. A dataset of 2,500 reviews was manually labeled based on two core UMUX-Lite indicators: perceived ease of use (P1) and perceived usefulness (P3). The labeling results were as follows: 35.1% irrelevant, 29.5% relevant to P1, 28.1% relevant to P3, and 7.4% relevant to both P1 and P3. The text underwent preprocessing, including stop word removal, tokenization, and stemming, before being analyzed via word cloud visualization, Term Frequency-Inverse Document Frequency (TF-IDF) weighting, and multiple classification algorithms (Naive Bayes, Random Forest, Support Vector Machine (SVM), Logistic Regression, K-Nearest Neighbors (K-NN), and Decision Tree). To mitigate class imbalance, the Synthetic Minority Oversampling Technique (SMOTE) was applied. Post-oversampling, all classifiers achieved accuracy rates above 80%. The results objectively indicate that users generally perceive the JMO application as easy to use and effective in meeting their needs, validating the proposed evaluation framework. The constructed corpus also serves as a valuable resource for future research on evaluating Indonesian-language mobile applications.]]>
Copyrights © 2026
