<![CDATA[This study aims to analyze visitor sentiment toward public facilities at Teras Samarinda based on user-generated reviews collected from digital platforms. The increasing number of online reviews provides valuable insights into visitor satisfaction; however, manual analysis is inefficient due to the large volume of data. Therefore, this research applies a text mining approach to automatically classify sentiments into positive, negative, and neutral categories. The dataset consists of 165 comments obtained from YouTube, representing visitor experiences and opinions. The preprocessing stage includes case folding, cleaning, tokenization, stopword removal, and stemming to ensure data quality. Subsequently, Term Frequency–Inverse Document Frequency is used to transform textual data into numerical features. The classification process is performed using the Naive Bayes algorithm. The dataset is divided into training and testing data to evaluate model performance using accuracy, precision, recall, and F1-score metrics. The results show that the model achieves an accuracy of 75.75%, indicating a relatively good performance in classifying sentiments. However, the model demonstrates limitations in distinguishing negative and neutral sentiments due to imbalanced data distribution. The findings reveal that most visitors express positive sentiment toward public facilities at Teras Samarinda, suggesting overall satisfaction. This study contributes to providing insights for improving facility quality and highlights the importance of handling imbalanced datasets in sentiment analysis.]]>
