Everybody plans vacations, and the first step in that process is to book a hotel. With the hospitality sector being so competitive, it’s critical to maintain best practices and stay on top of client demands and wants. They want individualized experiences, one-of-a-kind amenities, and a general sense of well-being on all levels. A consumer of a hotel recommendation system frequently encounters challenges in obtaining and fulfilling his or her wishes. Content-based filtering and collaborative filtering are two well-known strategies for creating a recommender system. Content-based filtering does not use human opinions to produce predictions, whereas collaborative filtering does, resulting in more accurate predictions. Collaborative filtering, on the other hand, cannot forecast objects that have never been rated by anyone. Both approaches can be merged with a hybrid methodology to cover the disadvantages of each approach while gaining the benefits of the other. This research employed Item-Item collaborative filtering (CF) and content-based filtering (CB) to calculate hotel similarity in our suggested method. It uses cosine similarity to calculate user similarity. For content-based filtering, natural language processing (NLP) is also employed. Our model employs a knowledge-based approach for Cold-User scenarios. Precision, recall and f1 used to evaluate the recommendation system.