Designment And Implementation Of Personalized Movie Recommendation System

With the popularization of Internet in daily life, a lot of garbage informationbrings us problems to finding the effective information. In order to solve the problemof information overload, the personalized recommendation system came into being. Itmakes good use of user behavior information to do data mining and informationfiltering, recommending products or items in which users might be interested.For film lovers, searching the Internet movie database for movies they like is notan easy issue. As a carrier of art, the movie has its own objective data, but alsocontains a large number of subject information, like comments and tags add by others.People often have reference to the opinions of others to do judgment. Therefore, thefilm personalized search&recommend system will provide users with personalizedservice of search and recommend.This paper introduces the background and the framework of personalizedrecommendation system at first, and then improves the traditional user-basedcollaborative filtering algorithm in two respects, one of which is adding a userclustering process before collaborative filtering and the other is adding a weight factorof user points in the system to the prediction formula. Meanwhile, the analysis, designand realization of the searching and recommending part of the system are alsoelaborated. The system is based on ASP.NET framework and Microsoft SQL Server2005database, while the algorithm is realized by mixed programming of C#andMatlab. The last part analyzes the time complexity and accuracy of the algorithmthrough experiment, proving that the efficiency and accuracy of recommendation isimproved.