Analisis Sentimen Ulasan Pada Google Review Di Sebuah Penginapan Menggunakan Algoritma Naive Bayes

This study aims to analyze customer sentiment towards Grand Jatra Hotel Pekanbaru on Google Review using the Naïve Bayes algorithm. Social media and online review platforms have become key sources of information for potential customers in making decisions. Using web scraping techniques, review data is collected and processed through preprocessing stages, including case folding, tokenizing, normalization, stopword removal, and stemming. The TF-IDF method is used to weight words before classifying sentiments into two categories: positive and negative. The Naïve Bayes model is then trained using training data and tested to evaluate its performance. The evaluation results show that the model achieves an accuracy rate of 98%, with precision 97% and recall 100% for the positive class at 92% each. These findings indicate that the Naïve Bayes algorithm can effectively classify customer sentiment toward hotel services and facilities. The results of this study are expected to provide insights for hotel management to improve service quality and marketing strategies based on customer sentiment analysis.