Web-based travel scheduling and booking has become one of the major commercial uses.Hotel booking websites use online rating and customer input to support the decision-making process of the client, but reviews provide a better insight into the hotel, but most travelers do not have time or patience to read all reviews. This research analyzes the ratings of hotels and provides information that may miss. The comments and metadata are crawled from the website and grouped according to some specific aspects into pre-defined categories.Here, we try to make efficient reviews sentiment analysis on “booking.com” hotel reviews and apply NLP to pre-processing of data. After that identifying the subjective information in text and classifying each piece of data as positive, negative and neutral response. This pre-processing data convert into vector and apply convolution neural network(CNN) algorithm on vector matrix and outcomes of CNN represent as pie-chart and bar-chart using Django library. This chart are define base on categories like room, food, cleanliness, service, staff, nature view, facilities.