A machine-learning-based recommender system for hotel reviews

With the rapid growth of online review platforms, selecting suitable accommodations has become increasingly challenging due to the overwhelming volume of user-generated content. This study presents a machine learning-based hotel recommendation system that processes Google Reviews for hotels in Perak, Malaysia. The system aims to enhance decision-making by recommending hotels based on user preferences, such as star ratings and sentiment features extracted from textual reviews. A total of 2,993 reviews from 58 hotels were collected using a Python-based scraper and pre-processed data to remove noise and tokenize content. Three machine learning models, namely Naïve Bayes, Random Forest, and Support Vector Machine, were evaluated using standard performance metrics. The SVM model achieved the highest accuracy (74.46%) and was selected for integration into a web-based application developed using Flask. The final system allows users to filter hotel recommendations by name, rating, and custom keywords, providing personalised suggestions and valuable insights. This project demonstrates the potential of machine learning to improve user experiences and decision-making in the hospitality industry.