The Research And Application Of Personalized Recommendation System Based On Collaborative Filtering Used In Mail Platform

Personalized recommendation system based on the user's interests,give users personalized recommendation,and personalized recommendation system has been widely used in e-commerce,social networking sites and other Internet application system,which won a lot of users' praise.For many years,send and receive email has been an important social communication in people's work and life.In recent years,with the advent of the mobile Internet era,the use of mail in the phone's applications is more and more frequently.However,as an essential social applications,the mailbox platform is still relatively lacking in recommended of personalized advertising.Currently,most advertising for the mailbox platform is the same.Most users are not interested or even disgusted with such ads,advertising effectiveness is greatly reduced,and the ads can not achieve the initial purpose.Based on this,this paper aims to achieve personalized advertising recommendation based on the mailbox platform.Based on the study of a large number of personalized recommendation algorithms,this paper presents a personalized recommendation algorithm based on the mailbox platform,combined with the characteristics of the mailbox platform,design the main framework and algorithm of the recommendation system.This paper proposes to use based on content and collaborative filtering hybrid recommendation algorithm,though extract and analysis the registration information,e-mail messages and e-mail exchanges information of the user of mailbox,access to the user's preferences,according to the user's preferences and the characteristics of goods,give the user initial recommendation.After the system running for a period of time,use the collaborative filtering recommendation algorithm,combined with the user similarity based on trust mechanism,calculate the similarity between users,get the user's nearest neighbor set and project prediction score,and finally get the recommended items.At the same time,according to the attribute of the item,this paper proposes a collaborative filtering algorithm based on item classification.The prediction of the project can be obtained by calculating the similarity of the project in the item class.To a certain extent,the sparsity of data is alleviated.The method proposed in this paper has achieved good results.