Recommendation systems have experienced a surge in popularity, with their adoption on the rise in various fields, including e-commerce, films, tourism, news, advertising, stock markets and social networks. This review paper focuses on machine learning approach utilized in hybrid recommendation systems aims to enhance the precision and variety of recommendations by integrating multiple recommendation methods. In addition, the paper introduces the methodology of the TF-IDF, which incorporates n-grams to capture the weight of the term more accurately by taking into account the frequency of the n-word sequence with taking into consideration on evaluating the appropriate weights of different features based on customer preferences and industry standards. Moreover, it incorporates sentiment polarity analysis using deep learning techniques, encryption techniques and matrix factorization for analysis of hotel data. Overall, the research papers cover a wide range of techniques and methods for analyzing and extracting valuable insights from hotel data extracted from TripAdvisor.com on Hospitality Sector