Design And Implemention Of Coupon Precision Marketing System Based On Web

With the development of the Internet,the traditional retail industry has continued suffering from sluggish performance under the impact of the e-commerce industry.Coupon is a popular marketing method used by retail companies.However,the delivery effect of traditional group coupon is decreasing,resulting in the continuous loss of user resources.Therefore,by using the coupon function of Wechat public platform which is widely used at present,this thesis designs and implements a coupon precision marketing system based on Web.And our proposed system has important significance and application value,which aims to help retail companies to achieve precision marketing of coupon through the implementation of user profile and coupon recommendation.In order to solve the existing management,promotion,maintenance and lack of relevance problems of traditional group coupon,this thesis designs and implements a coupon precision marketing system based on Web.And the functional modules of our proposed system include precise coupon module,user profile module,marketing activity management module,and data report center module.The implementation of the system Web platform is based on the MVC design pattern,and the key technologies include jQuery,Ajax and ThinkPHP.The front-end pages are used for displaying data,getting user operations and data visualization.The backstage is in charging of controlling the business logic of coupon management,operating database,calling WeChat public platform interfaces and interacting with the front-end.The database is a data storage warehouse that supports the data services of the entire system.And the test results show that each module is operating normally and stably.First,in order to implement the precision marketing of coupon,our system implemented user profile based on history sales data and user data of retail companies.The user profile generates multiple dimensions of labels for users by using statistical methods and K-means algorithms,it analysises consumer behavior of different types of users and provides an effective basis for retail companies to implement the precision marketing of coupon.Second,the coupon recommendation provides brand group data support for the system to achieve the precision marketing of coupon.In the progress of implementing the coupon recommendation,we first preprocessed the data.Then we used the TensorFlow deep learning framework to construct an Additional Denoising Autoencoder model,and the user labels are inputted into the model as side informations to optimize the entire algorithm from multiple aspects.Besides,the performance advantage of the collaborative filtering algorithm based on Additional Denoising Autoencoder was verified through rating predicet and Top-N recommendation experiment.