Hotel big data smart cache prototype

With the increase of clients and hotel resources in Amadeus Hotel, it becomes a challenge for Amadeus Hotel to provide personalized hotel information for different clients and to optimize the Hotel searching process for both client and Amadeus sides. Nowadays, recommendation systems are playing a very important role in the domain of recommendation services, specifically online searching and shopping process. It could provide matches between users and items in order to provide attractive information which are matching their preferences. To overcome the challenge of Amadeus Hotel, a recommendation system will take steps to optimize the search result and provides hotels that the client actually needs for Amadeus Hotel. This paper proposes a recommendation solution to classify hotel profile and traveler profile and making the recommendation for Amadeus Hotel shopping processes. This solution could recommends personalized hotels for given users. The proposed approach combines customer reviews analysis, content based filtering and collaborative filtering with matrix factorization and classification techniques to improve the performance and it is benefitting from Amadeus large-scale data and hotel rating system. In this paper, it describes how a better hotel recommendation system have been designed and built, how this recommendation system have been integrated with several feature information and systems. The evaluation results shows how this recommendation system could improve the process of Amadeus hotel shopping and user experiences.