Design Of Tariff Package Scheme Based On User's Internet Behavior

With the rapid development of the mobile Internet market,the existing packages mechanism of operators has become the shackles of the continued growth of the mobile Internet because of its high price.In this context,Internet companies began to cooperate with mobile operators to propose a flow-free package based on specific access content,such as Tencent King Card.The design of such flow-free package mechanism is bound to be the direction of future research.Based on the real data set provided by operators,this paper uses user portrait correlation method to subdivide user groups and set corresponding packages according to the characteristics of different user sub-groups.The pricing mechanism of packages is designed by combining machine learning algorithm and economic related models.The main research contents are as follows:1.Based on the real data set,the attributes of active users and accessed content are analyzed.The analysis of active users proves the feasibility of the current free-flow package.In the analysis of access content,this paper studies the user stickiness and traffic demand of access content based on clustering method.2.Based on the polymorphic model,the user's online behavior is modeled and analyzed.The user's continuous online behavior is characterized by the transfer matrix,and a factor model based on tensor decomposition is proposed to subdivide the user group.In this paper,the above method is applied to the actual data,and good segmentation results are achieved.3.A set pricing mechanism combining bilateral market model and SoftMax algorithm is proposed.In this paper,the utility functions of users,Internet manufacturers and mobile operators in monopoly and competition scenarios are designed respectively.Under the optimization objectives of social welfare optimization and platform profit optimization,the utility functions are solved quantitatively by combining SoftMax algorithm.In this paper,user portrait correlation method is used to subdivide the user groups,and the corresponding packages are designed for each subgroup according to the subgroup characteristics after subdivision.Subsequently,this paper uses the proposed pricing mechanism to price the packages designed in this paper,which can provide reasonable suggestions for mobile operators.This paper combines machine learning algorithm with economics to deal with practical problems,which has certain application value,and provides a new way of thinking for the follow-up study of operator packages mechanism.