| In recent years,the concept of data science and big data has been intensified by the media.Data is not limited to data generated by the Internet.Financial,medical,pharmaceutical,,government,retail,and other industries generate large amounts of data,and the influence of data is increasing day by day.People's daily behaviors are digitized,online shopping,communications,browsing news,search information and reviews are all recorded.We can rely on sufficient and inexpensive computing power to conduct in-depth studies of human behavior and better understand human beings.User portrait is equivalent to user information tagging and dataization,Companies collect and research consumer static data and dynamic behavior data,use user portrait to help the company determine the target user groups and meet the needs of various users.With the increase of e-commerce market,the major e-commerce sites and value of the user data drivers are listed as the company's values.Traditional user portraits rely too much on objective and past experience.Therefore,a user-friendly and powerful user portrait system is designed to make the user's analysis platform simplistic,intelligent,and turn data into actual business value.This paper base on the perspective of the e-commerce operations analysis staff,combines the stable user portrait scheme and architecture commonly used in the company's production environment,while referring to the emerging technologies and tools emerging in the big data environment,and proposes that the stable architecture be combined with emerging technologies to optimize each other.The user's portrait system consists of label system,indicator data,user view,aurora platform,work center function to be a sound software system.The front-end of the system uses the popular Bootstrap framework to make the layout of the page reasonable and clear.At the same time,advanced front-end technologies such as Ajax,jQuery,and Echarts are used to visualize the data.The back end uses a high-performance Django development framework,and combines the Hadoop distributed computing framework to store the source data by e-commerce users using E-commerce website,including static information data,dynamic behavior data.This system use the MySQL database to save system function management data and the user's attributes.Labels,demographic information,and business statistics are stored using the MongoDB non-relational database,thereby providing high-performance,highly reliable back-end storage capabilities to build a user-friendly portraiture system with high security and stability. |
