Research Of Collaborative Filtering Recommendation Algorithm

With the rapid development of e-commerce,customer-centered relationship replaces the relationship with commodity-centered,and customers become the main body of the electronic commerce enterprises,how to attract customers and provide customers with better service are two important tasks at present.So recommendation system has become an important method to solve the problems in business.The work of the dissertation is partly supported by the National Natural Science Foundation of China(No.61172072,61271308),Beijing Natural Science Foundation(No.4112045),and Research Fund for the Doctoral Program of Higher Education of China(No.20100009110002).The main work and contributions of this paper are summarized as follows:First of all,the recommendation system is briefly introduced,including the concept,composition,classification and recommendation methods in practice.Collaborative filtering algorithm is explained in detail,including the basic idea,the classification,the algorithm steps,the advantages,the evaluation index,the main problems and the solutions.Secondly,in order to find the nearest neighbor user set,the artificial immune system is introduced into the collaborative filtering algorithm,and the influence of the user preference on the recommendation is fully considered.The selection process of antigen and antibody in the immune system is similar to the process of finding nearest neighbors set for the target users.And we combine with the user preference factor,thus put forward the artificial immune recommendation algorithm based on user preferences.But considering the algorithm did not solve the cold start problem,on the basis of fully excavate the user's personal information,we get the final artificial immune algorithm based user preference and user registration information.Finally,in order to verify the validity of the algorithm,we use the MovieLens data sets to take experiments,experimental results show that the final algorithm on the results is optimal,obviously the effect of collaborative filtering algorithm significantly improved.