Design And Implementation Of The Hotel Reservation And Occupancy System Based On The Personalized Recommendation

The rapid development of Internet technology has not only brought a lot of convenience to people's life, but also changed the way people live. More and more people prefer booking hotels, tickets, delivery and so on online. Online booking has become a trend, and the traditional business pattern and the management mode of hotel have been unable to meet the requirement that the trend brings. Therefore, to develope a hotel reservation and occupancy management system is of important significance.In the age of Web 2.0, more and more information are exposed to people with the explosive growth of information, which brings a serious problem named information overload. Personalized recommendation system is one of the most effective way to solve the problem, which can dig up user's interest and provide users with personalized services. A large hotel reservation platform, with the increasing number of the hotels on platform, will also face the problem of information overload. When people start booking hotel on the platform, they need to spend plenty of time to find the information of their interest. That is why it is necessary to develop the function of personalized recommendation in hotel reservation platform.This thesis firstly analyses the deficiency of the traditional hotel reservation and occupancy management systems, and puts forward improvement scheme for these deficiencies. At last, we use Spring and Hibernate framework to develop the hotel reservation and occupancy management system of B/S structrue, and add the personalized recommendation feature to the system.This system adopts the REST style to design the interface of server. The server is merely responsible for providing all sorts of REST style interface and not in charage of the webpage jumping of clients. This structure can completely separate server and client. In this architecture, the web pages and applications are able to access the same set of Restful interface, which can lighten the workload of background development. For the hotel recommendation system, we use user-based and item-based hybrid collaborative filtering algorithm to do recommendation for users, and in order to mitigate the effects of sparsity of hotel reviewing data on the recommendation result, this thesis proposes the method of using the amount of user's booking to evaluate the user ratings. Finally, in order to generate the recommendation model quickly, the thesis implements the recommendation algorithm in Hadoop distributed environment, and realizes the distributed processing of data. There are two situations that use recommendation function in this system.The first situation is recommending hotels for users in the form of list directly. The other one is the scene that when users search for hotels, the system will sort the results by the value of interest degree, and then return the sorted results.