<![CDATA[Online reviews on digital platforms significantly influence tourist decision-making. This study performs sentiment analysis on visitor reviews of the Guci Hot Spring tourist destination using machine learning algorithms. A total of 1,127 reviews were collected from Google Maps and preprocessed through cleaning, case folding, normalization, tokenization, stop-word removal, and duplicate elimination. Sentiment labelling was performed using a lexicon-based approach, classifying reviews into positive, negative, and neutral categories. The dataset was split into 80% for training and 20% for testing. Two algorithms were compared: Support Vector Machine (SVM) and Decision Tree (DT). Performance evaluation employed accuracy, precision, recall, and F1-score metrics. Results show that SVM outperforms Decision Tree, achieving 69.9% accuracy and 0.680 macro-average F1-score, compared to Decision Tree's 64.2% accuracy and 0.622 macro-average F1-score. These findings demonstrate that SVM is more effective in handling high-dimensional text data, particularly in distinguishing ambiguous neutral-class reviews. This research provides valuable insights for tourism management to improve service quality based on visitor sentiment patterns.]]>
Copyrights © 2026
