<![CDATA[The development of tourism in Indonesia offers vast business opportunities for tourism stakeholders, especially in Yogyakarta, which is rich in natural tourist destinations. However, visitors often face difficulties in selecting suitable destinations due to limited information. This study aims to develop a tourism recommendation system in Yogyakarta by utilizing the Naïve Bayes Multinomial algorithm to improve the accuracy of recommendations and the overall tourism experience. The system leverages word frequency information from textual data to generate better recommendations. Data from 40 tourist destinations, including descriptions, locations, ratings, categories, and facilities, is used in this study. The model's results show an accuracy of 81.2%, precision of 62.5%, recall of 62.5%, and an F1 score of 62.5%. Testing with the keyword "photo spot" yields an accuracy of 50%, precision of 25%, recall of 50%, and an F1 score of 33.33%, while the keyword "natural scenery" achieves 100% accuracy, 100% precision, 100% recall, and an F1 score of 100%. These results demonstrate that the system can provide accurate and useful tourism recommendations for users.]]>
Copyrights © 2025
