Hotel Reservation System Based On Personalized Room Type Recommendation Of Deep Learning

With the deep integration of the Internet and people’s life style,online travel agencies use the network platform to change the traditional tourism marketing mode,transmit offline hotel information and route information on the network,accelerate the dissemination and acquisition of hotel related information,and provide great convenience for users to book hotels.However,with the rapid growth of hotel information data,it takes a long time for users to find their preferred hotel types and hotels when booking hotels.Therefore,how to provide users with the hotel room type and the accurate push of the hotel is an important problem that the existing hotel reservation platform urgently needs to solve,and it is also the development trend of the future hotel reservation platform.In this thesis,a hotel reservation system based on personalized room type recommendation of deep learning is developed.In the system design,we first develop the system prototype according to the requirements,and then continuously change and perfect the system prototype until the system meets the user’s requirements.The main functions of this system include personalized recommendation,personal center,reservation management,hotel management,room management,system management and other functional modules.In this thesis,the DeepFM+XGBoost fusion model is built with Python language through the TensorFlow and XGBoost open source library.And we use the advantages of XGBoost for low-order feature extraction and DeepFM for low-order feature and high-order feature interaction to mine the user’s interests,hobbies,historical order behavior data,obtain various data features that potentially affect the user’s behavior,and train the model to get the best fusion model.In this thesis,the best fusion model is applied in the personalized recommendation function module to predict,and the TopK of the predicted result table is returned to the user,so as to realize the accurate recommendation of what kind of hotel room the user want to order,and make up for the shortcomings of the existing hotel booking platform,such as Ctrip,Qunar,etc.,that only recommend the most popular hotels,and all the recommended hotels are the same,but it did not recommend the most preferred hotel room type and hotel based on people.This thesis implements a system that has personalized recommendation,personal center,reservation management,hotel management,room management,system management and other functional modules.And the DeepFM+XGBoost fusion model is used for recommendation to get better recommendation accuracy.System function test is passed,and system performance is good.The system can basically provide hotel reservation,hotel room type and hotel personalized recommendation services.