Design And Implementation Of Book Recommendation System Based On Improved UserCF Algorithm

With the continuous deepening of the third industrial revolution, information technology has achieved explosive progresses in the global. Under the cover of information age, its results can be seen everywhere in society. Society is developing towards digitalization and informatization. Users find it is difficult to pick out the right books they want from a huge number of books in the school library. Libraries are faced with an embarrassed situation, which is "many books in the libraries to meet the needs of users but users cannot find them quickly".Therefore, in order to better help these users to find the books they want to find, Personalized Recommendation System has been an important research field for the university library. This paper includes the following three parts:(1) Architectural design of book recommendation system. Book recommendation system architecture comprises the input and the output, the recommended method of three parts. The framework of this system is a combination of MVC architecture, Hibernate framework and struts framework.(2) The detailed design of book recommendation system. System functions include system management module(user management, books management, books query), order processing module, personalized recommendation module(offline and online recommendation module), and other functional modules.(3) Achieving a personalized book recommendation service by using CF algorithm. It firstly calculates the similarity data of all kinds of user in the library, then sorts the similarity of users to look for the nearest neighbor to get the most similar data, and finally realizes the personalized book recommendation service according to the most relevant records of borrowed books. By the approach of a series of site tests and investigation, the system can usually satisfy the users’ demand as well as achieving good effects.