示例：电商平台下用户画像的设计与分析

　　摘 要
　　大数据是数量非常巨大的复杂的半结构化或者非结构化的数据集。
　　随着时代的发展，越来越多的数据产生，用传统的数据库管理方式，无论是从数据的储存，获取，或者查找等方面都已经无法满足要求了。随着大数据的来临，人们有了解决大量数据的处理，储存等能力后，人们为了将数据细化，将数据分行业，类型的分了好多种类。在这里我们所研究的是其中的一个种类：用户画像。
　　用户画像核心价值在于了解用户，猜测用户对产品的需求或者潜在需求，精细化的定位人群特征，挖掘潜在的用户群体，为媒体网站、广告主、企业及广告公司充分认知群体用户的差异化特征，根据族群的差异化特征，帮助客户找到营销机会、运营方向，全面提高客户的核心影响力。在电商行业中。用户画像可以分析客户的使用习惯、喜好、一系列的购买行为，以及周边的人群的身份、属性、年龄等，它对我们的商务行为、对营销的判断会有一定的借鉴。
　　本文通过使用Hadoop技术，结合Hive，Java，JSP和HTML等编程语言，设计并完成了一个具有简单的电商平台下的用户画像。本设计依赖Easyui,Echarts,JfreeChar框架设计出了简洁漂亮的前端界面，使用Hive 进行数据分析与产生研究结果。本系统主要包括,系统管理：分别从用户管理，角色管理，菜单管理等方面对整个系统进行权限管理。用户行为：分别从跳出率，忠诚度，活跃度判断用户在某一天或者某一段时间的整体状况。访客分析：分别从地域分布，速度分布，客户端环境等对某个地区的整体环境和地区消费情况进行产品销售。
　　关键词：用户画像； Hadoop； Hive
　　ABSTRACT
　　Big data is a very large number of complex semi structured or unstructured data sets.
　　With the development of the times, more and more data are produced, with the traditional database management, whether it is from the data storage, access, or find and so on have been unable to meet the requirements of the. With the advent of big data, people have to deal with a large number of data processing, storage and other capabilities, people in order to refine the data, the data points industry, a lot of types of points. What we are studying here is one of the categories: the user portrait.
　　User portrait core value is to understand the user, users of the product demand or potential demand forecast, fine positioning population characteristics, mining the potential user groups, web media, advertisers, and advertising companies fully cognitive differentiation characteristics of the user groups, according to the features of the ethnic differences to help customers find opportunities for marketing, operations, and comprehensively improve the customer's core impact. In the electricity business industry. User portrait can analyze customer use habits, preferences, a series of purchase behavior, and peripheral populations of identity, attributes, age, it to our business behavior, the judgment of the marketing will have a certain reference.
　　This article through the use of Hadoop technology, combined with Hive, Java, JSP and HTML programming languages, designed and completed a simple business platform with a user portrait. The design of Easyui, Echarts, JfreeChar framework designed a simple and beautiful front interface, using Hive for data analysis and research results. The system mainly includes the system management: from the user management, role management, menu management and other aspects of the entire system to carry out the rights management. User behavior: respectively, from the jump out rate, loyalty, activity to judge the overall situation of the user in a day or a certain period of time. Visitor analysis: from the geographical distribution, speed distribution, client environment, such as the overall environment of a region and regional consumer sales.
　　Key words：User Profile;Hadoop; Hive

请下载文档查看详情资料