A novel approach for hotel recommendation system based on modified KNN and naive bayes algorithm.

For the purpose of assisting customers in selecting the most appropriate hotel within their budgetary constraints, close to convenient amenities such cafes and internet cafes. This study helps in simplifying the process of selecting the most suitable hotel, ensuring a pleasant stay with all conveniences. The hotel datasets in the kaggle dataset were used to acquire a total of 730 samples. The modified KNN method and the naïve bayes algorithm were employed for this purpose, and their performance was evaluated and compared. The value of G-force was stable at 80%. When compared to the results obtained by the naive bayes algorithm (80.1% accuracy, 83.39 percent precision, 84.2 percent sensitivity, and 86.2 percent specificity), the results obtained by the modified collaborative filtering were, respectively, 87.4 percent, 88.2 percent, 89.2 percent, and 93.3 percent. More than 730 samples were used, and the findings were statistically significant at the 0.162 level, with a pretest power of 80%. The results show that the new technique is more accurate than the naïve bayes algorithm. [ABSTRACT FROM AUTHOR]