This study analyzes public sentiment toward the South Sumatra State Museum based on review data collected from Google Maps. A total of 1,200 reviews were gathered and manually labeled into positive and negative sentiment categories. The collected data underwent text preprocessing and feature extraction using the Term Frequency-Inverse Document Frequency (TF-IDF) technique. Model performance was evaluated using three classification algorithms—Support Vector Machine (SVM), Random Forest, and Naïve Bayes—across data split scenarios of 80:20, 60:40, and 50:50. The performance of the models was measured using four evaluation metrics: accuracy, precision, recall, and F1-score. The evaluation results show that under the best scenario for each method, SVM (60:40) achieved an accuracy of 85.4%, precision of 86.6%, recall of 94.1%, and an F1-score of 90.5%; Random Forest (80:20) recorded an accuracy of 81.6%, precision of 89%, recall of 84.8%, and an F1-score of 86.5%; while Naïve Bayes (50:50) reached an accuracy of 80.1%, precision of 79.8%, recall of 96.7%, and an F1-score of 87.4%. The contribution of this study lies in providing an experimental, evidence-based comparative analysis to assist museum management in understanding visitor perceptions objectively and measurably.
Copyrights © 2026