节点文献
移动公司客户画像系统设计与实现
Research on Design and Implementation of Customer Portrait System in China Mobile
【作者】 杨明;
【作者基本信息】 电子科技大学 , 工程硕士(专业学位), 2020, 硕士
【摘要】 随着大数据技术的发展,利用大数据技术搜集客户信息,并整合为客户画像,成为行业内的重要策略手段。这一应用对深入了解客户需求,设计客户真正需求的产品具有重要的意义。因此,本文研究移动公司的客户画像系统,利用大数据技术整合相关数据,形成客户的画像标签,并为业务人员提供数据挖掘功能,这对提高移动公司的竞争力、增强客户黏性具有重要的价值。本文主要对移动公司的精准营销画像系统进行研究。首先,本文对画像系统的需求进行分析。目前,画像系统需要实现客户数据的整合,从各个业务系统中将客户相关的数据通过ETL工具进行汇总以及清洗后,统一存储到大数据平台。同时业务部门人员根据业务规则设计客户的标签,由大数据评价计算客户的标签。同时,画像系统还需提供数据挖掘功能,支持业务人员对客户数据的分析与挖掘并对外提供数据的访问接口。其次,本文在需求分析的基础上,对系统进行设计,包括对系统的整体逻辑、性能架构以及数据架构进线设计。本系统基于大数据平台用于存储各种画像基础数据,通过对基础数据进行汇总、规则校验以及建模,形成画像系统的标签数据处理结构。在对外部系统的接口上,设计了接口规范,最后对可视化交互需求进行了原型设计。第三,本文详细分析了数据清洗以及数据存储的实现方式。本系统基于Hadoop集群上的Hive数据仓库用于存储各种画像基础数据,基于Restful技术实现外部系统的接口,最后采用Echarts框架实现系统的交互式界面。第四,本文详细研究了画像标签的设计与实现。同时,详细研究了画像系统中数据挖掘功能的实现,采用PMML接口,对外提供机器学习功能,让外部系统能够通过画像系统调用机器学习模型,进一步扩展了系统的可用性。本文的研究为移动公司建设精准营销系统,完善客户画像体系具有一定的参考,也为其他类似公司建设画像系统提供了借鉴。
【Abstract】 With the development of big data technology,using big data technology to collect customer information and integrate it into customer portrait has become an important strategic means in the industry.This application is of great significance for us to understand the customers’ needs and design the products they really need.Therefore,this paper studies the customer portrait system of mobile companies,uses big data technology to integrate relevant data,forms customer portrait labels,and provides data mining functions for business personnel,which is of great value to improve the competitiveness of mobile companies and enhance customer stickiness.This paper mainly studies the precision marketing portrait system of mobile companies.First of all,this paper analyzes the requirements of the portrait system.At present,the image system needs to realize the integration of customer data.At the same time,the portrait system also needs to provide data mining function,support business personnel to analyze and mine customer data,and provide external data access interface.Secondly,on the basis of requirement analysis,this paper designs the system,including the overall logic,performance architecture and data architecture design.Based on the big data platform,this system is used to store all kinds of basic data of portrait.Third,this paper analyzes the implementation of data cleaning and data storage in detail.This system is based on hive data warehouse on Hadoop cluster,which is used to store all kinds of basic data of portrait.It realizes the interface of external system based on restful technology.Finally,it uses echarts framework to realize the interactive interface of the system.Fourthly,this paper studies the design and implementation of portrait label in detail.At the same time,the realization of the data mining function in the portrait system is studied in detail.The research of this paper has some reference for mobile companies to build precision marketing system and improve customer portrait system,and also provides reference for other similar companies to build portrait system.
【Key words】 mobile communication; user profile; tag system; clustering; association rules;
- 【网络出版投稿人】 电子科技大学 【网络出版年期】2021年 01期
- 【分类号】TP311.13
- 【被引频次】3
- 【下载频次】277