基于Web数据的港口客户分类及商务决策系统研究

发布时间：2018-03-31 00:25

本文选题：数据挖掘　切入点：客户分类　出处：《北京交通大学》2017年硕士论文

【摘要】：港口作为货物集散和贸易往来的场所,对一个国家或地区是非常重要的运输枢纽。随着港口服务市场近年来供求关系的转变,港口企业为谋求更长远的发展,试图通过传统商业智能技术辅助其决策。然而由于传统商业智能技术着重于分析企业内部的结构化数据,对外部的非结构化数据处理存在较大的局限性,导致其最终得到的应用效果并不理想。针对这一问题,本文提出了一种基于Web数据的港口客户分类方法,通过对企业外部非结构化数据进行采集和处理,有针对性的提取港口客户相关信息,进而运用文本分类、神经网络等理论和技术,以期使用外部数据结合港口内部数据对港口客户进行全面立体的刻画。内外部数据结合的大数据技术的综合应用克服了内部数据在港口客户分类管理中存在迟滞、片面、失真的问题,使得企业更加及时、全面、系统地了解客户真实情况,进而帮助港口企业更好的了解客户以进行有效的商务决策。本文以广州某港口企业信息化建设现状为研究背景,采用基于Web数据的港口客户数据结合传统内部数据的方法进行客户分类,以港口企业客户管理决策为目的,在研究客户管理理论、文本分类、客户分类方法的基础上做了如下研究:第一,通过网络爬虫技术获取港口客户外部数据并转化成分类指标,然后结合传统客户分类内部指标建立全面的港口客户分类指标体系;第二,基于内外部数据结合建立全面的指标体系搭建港口客户分类模型;第三,结合客户管理理论及港口业务流程,对内外部数据结合的客户分类结果进行针对性的商务决策支持;最后,利用分布式计算技术以及J2EE技术实现了该港口商务决策系统的设计。
[Abstract]:As a place for cargo distribution and trade, port is a very important transportation hub for a country or region. With the change of supply and demand relationship in port service market in recent years, port enterprises seek for more long-term development. This paper attempts to assist the decision making by the traditional business intelligence technology. However, because the traditional business intelligence technology focuses on the analysis of the internal structured data, it has great limitations to the external unstructured data processing. In view of this problem, this paper puts forward a method of port customer classification based on Web data, which collects and processes external unstructured data. The relevant information of port customers is extracted pertinently, and then the theory and technology of text classification, neural network and so on are used. The comprehensive application of big data technology combined with external data and port internal data can overcome the lag and one-sided of internal data in the classification management of port customers. The problem of distortion makes the enterprise more timely, comprehensive and systematic understanding of the customer's real situation, In order to help port enterprises better understand customers in order to make effective business decisions, this paper takes the present situation of information construction of a port enterprise in Guangzhou as the research background. This paper adopts the method of combining port customer data based on Web data with traditional internal data to classify customers. The purpose of this paper is to study the theory of customer management and text classification. On the basis of the customer classification method, the following research is done: first, through the web crawler technology to obtain port customer external data and transform into classification indicators, and then combined with the traditional customer classification internal indicators to establish a comprehensive port customer classification index system; Second, based on the combination of internal and external data to establish a comprehensive index system to build a port customer classification model; third, combined with customer management theory and port business processes, Finally, the design of the port business decision system is realized by using distributed computing technology and J2EE technology.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP311.52;F552.6;F274

【参考文献】