In the modern era, the rapid growth of the internet has made finding information about hotels much easier for tourists. As online reviews impacts a more substantial part in prompting consumer choices, it has become crucial for businesses, particularly in the hospitality sector, to analyze user sentiment. The internet has contributed a significant role in upholding the quality of hotels. The expansion of the hospitality industry in Bangladesh and West Bengal, India, has led to a surge in online hotel reviews. Analyzing these reviews is crucial for hotels to understand customer experiences and enrich their services. This study recommends a machine learning method for analysing sentiment on Bengali text-based hotel reviews. A dataset of Bengali hotel reviews was created, and natural language processing (NLP) techniques were applied to preprocess the data. A number of machine learning models, including Naive Bayes (NB), Support Vector Machines (SVMs), XGBoost, Decision Tree (DT), and Random Forest (RF), were evaluated for their effectiveness in classifying sentiments. The TF-IDF feature extraction technique was applied to convert the textual data into numerical representations. These models were evaluated using different classification metrics. Among the various models, RF achieved the best accuracy of 82% in sentiment classification. This research adds to the expanding area of Bengali NLP and establishes a basis for future studies in sentiment analysis and customer feedback assessment in the Bengali language.