<![CDATA[Digital transformation has become a strategic requirement for micro, small, and medium enterprises (MSMEs) in Indonesia, particularly in the adoption of marketplaces, QRIS-based payments, digital promotion, and platform-based business operations. This study analyzes public perception of MSME digital transformation using 5,751 YouTube comments collected from January to May 2026 through selected keywords related to MSME digitalization, including marketplace adoption, Shopee UMKM, Tokopedia UMKM, QRIS, and online business practices. The research applies a machine learning-based sentiment analysis pipeline consisting of text preprocessing, TF-IDF feature extraction, classification using Logistic Regression, Support Vector Machine, and Random Forest, and performance comparison with and without SMOTE. The evaluation uses accuracy, macro F1-score, confusion matrix, monthly distribution, keyword frequency, and word cloud visualization. The findings indicate that public discussion is dominated by neutral comments, while positive expressions highlight usefulness, ease, and marketplace opportunities. Negative comments are mainly associated with technical problems, application errors, costs, and platform difficulties. Random Forest without SMOTE achieved the highest accuracy of 98.94%, while SVM with SMOTE obtained the best macro F1-score of 83.77%, showing a better balance in recognizing minority sentiment classes. The study concludes that YouTube comments can function as a useful source of digital social sensing to understand public perception and to support evidence-based MSME digital transformation strategies.]]>
Copyrights © 2026
