The Design And Implementation Of The Iron Tower Online Bookstore Personalized Recommendation System

With the rapid development of Internet, web-based online bookstore came into being. The change from bookstores to online bookstores reduces labor costs and cost of sales for operators and attracts the customers from different regions. But it also weakened the interaction between the users and the websites, which is not conducive to the vendors to recommend books for users and promote sales. At the same time, various categories of books and the users’aimless browsing will greatly increase the difficulty in knowing about the users’preference for books. So what plays a crucial role in the book sales of website is how analyze and know about the user’s interests and providing a faster and better recommendation of their preferred books for them.In this paper, through the analysis of the development background on personalized recommendation and research on current situation, the Tower online bookstore’s analysis of the demand on personalized recommendation has been investigated in detail, and the implementation of system has been designed in outline and in detail. The SSH structure will be used in the system, and ExtJS will be used as the presentation layer to render technology for the realization of the book recommendation and management functions. The book recommendation are divided into real-time recommendation and off-line recommendation, in which collaborative filtering and association rule mining algorithm are used to perform different functions. And in the real-time part, the system will label the books with the collection of the key words at first, and then establish the interest model for each visitor. At the same time, a collection of recommended books are generated through the collaborative filtering algorithm based on the model for each user. Part of the system of the off-line recommendation will be daily mine the information of the web log and the sales record with the association rules to find the collection of the recommended books which meet the requirements.The development of this system will play the role of e-Purchasing Guide in providing users with different forms of functions in book recommendation and with more opportunities for book selection.