APP Personalization Recommendation Based On Collaborative Filtering Algorithm

The advent of the mobile phone era and the rise of the mobile applications(referred to as App)industry,make telecom operators launched APP products.The use of APP brings flux income to operators,but the market share of its products still needs to be improved.At the same time,the user’s personalized product requirements make the operator to adjust the marketing policy.Therefore,how to develop personalized recommendations for user preferences,is of great significance to operator.This paper is based on collaborative filtering algorithm to recommend the telecommunications industry’s own brand and high flux consumption APP products.First,analyze the current situation of the APP market,to find the factors that affect the user use APP,and get the list of products to be recommended.Then use factor analysis and logical regression algorithm to determine the APP score and user preferences,and according to user preferences to adjust the score.Next use the clustering algorithm to solve the sparseness of the user-project scoring matrix.Finally,make recommenddations to different type of users.Based on the behavior data of telecom users,this paper makes an individual recommendation on the APP products.The results show that it is effective and feasible to use the collaborative filtering algorithm to personalized recommend the APP products.The result of this research provide decision foundation for telecom operators to develop personalized recommendation program,improve their own brand market share while increasing revenue and improve user stickiness.