Sentiment Analysis of Students on Campus Facilities and Infrastructure Using the Naïve Bayes Classifier Method (Case Study STMIK Kaputama)

Campus facilities and infrastructure play an important role in supporting the quality of learning. STMIK Kaputama faces challenges in maintaining the quality of its facilities as the number of students increases. This study applies sentiment analysis to student comments regarding classrooms, laboratories, libraries, restrooms, parking, and internet access. The method used is the Naïve Bayes Classifier with TF-IDF weighting and text preprocessing, following the CRISP-DM framework. The results show an accuracy of 73%, with the best performance in the positive class with precision 0.72; recall 0.97; F1-score 0.82, while the negative class with precision 0.79; recall 0.38; F1-score 0.51 and the neutral class was not detected. These findings indicate that the model tends to be dominant in positive sentiment but is still weak in distinguishing between negative and neutral comments.