The rapid growth of tourism in Ambon City has increased competition among accommodations, necessitating data-driven performance evaluations. Prospective tourists often struggle with unstructured online reviews, while hotel management requires precise insights for improvement. This study aims to systematically classify hotel performance in Ambon City using the Naïve Bayes Algorithm based on reviews from platforms like Agoda and TripAdvisor. Adopting a descriptive quantitative methodology, the study processes and labels performance data as "Good," "Poor," or "Very Good." Findings demonstrate that the Naïve Bayes model is highly effective, achieving 91% accuracy. Evaluation via a Confusion Matrix confirms the model's reliability in predicting majority categories, proving that ratings and reviews are strong satisfaction predictors. While the model faces minor challenges with the "Poor" minority category due to limited data, the study provides strategic value. It offers management guidance for targeted improvements and helps tourists make informed decisions, ultimately enhancing the competitiveness of Ambon’s hospitality industry
Copyrights © 2026