<![CDATA[This study aims to classify user sentiment toward an ethics-based consumption application using the Multinomial Naïve Bayes algorithm. The application examined contains social and moral content, often provoking complex opinion expressions. A total of 2,000 user reviews were collected from Google Play Store using web scraping and processed through a series of text preprocessing steps: case folding, cleansing, tokenizing, stopword removal, and stemming. The data were converted into numerical form using the Term Frequency–Inverse Document Frequency (TF-IDF) method and labeled into three sentiment categories: positive, neutral, and negative. The evaluation results show that the model achieved a precision of 92%, recall of 100%, and an f1-score of 96% for positive sentiment. However, the model underperformed in recognizing neutral and negative sentiments due to class imbalance. This study contributes to understanding the limitations of probabilistic classification models in handling imbalanced public opinion in socially driven digital spaces.]]>
Copyrights © 2025
