Directional Push Management System For Marketing Activities

With the rapid development of the mobile Internet,the handling of operator marketing activities has changed from the traditional offline to online,but the existing marketing activities are recommended and published.The traditional marketing method does not combine the characteristics of user consumption and the characteristics of marketing activities to make personalized recommendations.The recommended channels for activities are too single and the marketing results are irrational.In addition,operators are hierarchically managed,and the business needs of operators in different cities and cities are different.Traditional marketing activities are released for independent development,which increases the pressure on operators and makes it difficult to form a unified management.In view of the inconvenience of the online marketing activities of operators and the blindness of marketing objects,this thesis designs different recommendation models for the package business and traffic value-added services of marketing activities,and develops marketing campaign directional push based on JavaEE enterprise application platform.The management system applies the recommendation model to the system.The specific work is as follows:(1)The system adopts the B/S architecture and implements the separation of the front end and the background based on the MVC design pattern.The front end is mainly for the display of the interface and the acquisition of user operation data.The background mainly implements business logic control,database operation,SMS message platform interface call and front-end interaction.The specific function modules include a user management module,an activity management module,a tariff management module,an activity review module,a recommendation management module,and a push path management module.Realize the convenient and flexible release of marketing activities,match the user's consumption behavior,and implement push in different ways.(2)For the package business,this paper extracts the user consumption behavior characteristics and package activity characteristics,and combines the extracted features to design a hybrid recommendation algorithm based on content and user collaborative filtering.Generate a set of recommended users based on business rule calculations by integrating user collections based on package content and user collaborative filtering(3)For traffic value-added services,this paper proposes a hybrid recommendation algorithm based on decision tree classification and content.The algorithm is divided into three steps: the first step is based on the user traffic consumption behavior,and the decision tree model is used to predict the category in which the user will handle the value-added service;the second step is based on the content of the traffic value-added service,and the calculation is most similar to the new online value-added service.The service category,the users in this category are the potential user sets of the new online value-added services;the third step is to calculate the potential user sets according to the business rules,and generate a recommended user set.This paper based on the real data set of an operator to verify and compare the algorithm,and prove the feasibility of the marketing activity recommendation algorithm.