Research And Design On The Collaborative Filtering Film Recommendation System

Recommended system is one of the effective tools to solve the information overload,it's a strong guarantee of customer loyalty strategy in the Internet competition.Traditional recommendation is based on the popularity,sales,advertising and other means,more and more can not provide users with personalized,personalized recommendation services.In this view,this paper designs a recommendation system based on collaborative filtering user.This system is mainly used in the field of film,which effectively solves the closed loop of film recommendation and high score recommendation in the film recommendation.And improve,improve the accuracy and efficiency of the recommended results.According to the purpose of the recommendation system,this paper proposes the implementation of the recommendation system by using the technology of Web network framework and Spark machine learning algorithm.Firstly,this paper analyzes the design requirement of the recommender system,analyzes the practicability,rationality and utility of the proposed film,and chooses the web design and development mode reasonably,and chooses the SpringMVC and Mybatis framework to integrate and develop the system performance.Secondly,the collaborative filtering recommendation algorithm is studied,and the collaborative filtering recommendation algorithm is mainly based on the proposed model of proximity and dimensionality reduction,this paper studies the recommended algorithms,compares the differences between them,proposes the problem of matrix sparsity and content limitation in neighboring algorithms,chooses the reduced dimension algorithm to solve the problem,and proposes the improved method for the cross least squares in the descending dimension algorithm.Thirdly,using the Moivelens data set as the original data,the algorithm MATLAB simulation of the data,according to the data analysis algorithm merits and demerits,the analysis algorithm realizes the rationality and chooses the optimal algorithm.Finally,through the Java and Spark development language to complete the network programming modules,the design of three modules of the system,the system required to choose the technology,complete the web framework Springmvc and transaction framework Mybatis integration,database table Structure Associationdesign,the module function flow design.In this paper,the film recommendation system,through the research recommendation algorithm,selects the best algorithm,the frame integration development completes each function module.HTML page structure to achieve the division,table structure analysis data,user login module,user query,user rating,user history and referral module,complete the module development,start the service after the completion of system testing,to achieve collaborative filtering film personalized recommendations,the results of the HTML display data.