Book Sales And Recommendation System Based On Collaborative Filtering

Personalized recommendation has been a key part for the modern electric business system. Personalized recommendation can enhance the successful buying probability significantly and convenience users’ buying behaviors. In order to obtain a personalized recommendation system, in this paper, we introduce a novel book selling system with a personalized recommendation function based on the collaborative filtering technique.In this thesis we first introduce the research background and objectives of this project. Specifically, E-commerce has profoundly changed the way people live and the business running methods. E-commerce system can successfully build up a convenient bridge between enterprises and ordinary users. Users can use the online shopping system to easily buy what they want. Operators can use the e-commerce system to sale merchandises easily. In this project, we design and develop a book sale and recommendation system based on collaborative filtering. In the introduction section,we also summarize the related work systematically from different angles, e.g.,e-commerce systems, recommendation systems, data mining and information systems research.We further conduct detailed system analysis based on the previous research results. On the whole, the system is divided into two parts, the first part is for the ordinary customer oriented front-end functions, and the second part is for the administrator back-end functionality. Front-end part contains several important functions including personal basic information management, personalized book recommendation, shopping cart management, order management, and user feedback functions. Back-end part is developed for administrators, it contains product management function, special offers management, order management, user management, as well as Super Administrator.Based on the results of the need analysis, we conduct system design and study on key algorithm. During the process of detailed design, the data flow graph of each function module is designed, which straightens the logic of each functional module.During the detailed design stage, the database structure is designed. We also describe the content of each database table in detail. Finally, we study the book recommendation algorithm based on collaborative filtering technique. The book recommendation algorithm is implemented through mining collective intelligence.In the process of system development and deployment, we first describe the overall program framework. This project adopts the MVC as the application architecture, which separate the view layer, model layer and control layer strictly. This separated system structure can greatly enhance system stability, reliability andscalability. We also introduce the method of system deployment, we use the My SQL as the database management system and Tomcat as application server.Finally, we summarize the full thesis briefly and give the possible improvement direction for this system in the future. We also highlight the importance of the mobile Internet. In the future, we will develop special APP for the book sale system running in mobile client.