ETL在期货CRM数据中心中的应用研究

发布时间：2018-05-13 13:40

本文选题：CRM + 数据中心　；参考：《浙江工业大学》2012年硕士论文

【摘要】：近几年随着期货行业的不断发展，期货业对IT技术的依赖程度越来越高，，柜台交易系统、清算系统、资产管理系统、办公自动化系统、财务系统都相继建成，但目前应用仍以业务运行为主，各个系统间信息缺乏有效的共享手段，形成了大量的“信息孤岛”，缺乏强大的业务支持、管理控制、业务分析和决策支持功能。期货公司为了谋求更大的发展，都在积极变革思维，从外部的扩张转向到内部的管理，从以“业务为导向”转向到以“客户为中心”，从“佣金价格竞争”转向到“服务质量竞争”，期货公司要在激烈的市场竞争中脱颖而出，必须重视公司的核心价值。对于期货公司来说，客户是真正的价值所在，客户关系管理（CRM）作为一种提升客户满意度、提高客户忠诚度、深层次开发客户价值的有效应用受到国内期货公司越来越多的重视。建设以数据中心为基础的CRM系统成为当务之急。对于CRM系统而言，最重要的是全面丰富准确的客户数据，为了满足CRM的各种应用，建立统一的客户视图，建设分析型企业级数据中心是基础。在数据中心建设过程中ETL（Extract、Transform and Load，数据抽取、转换和加载）和数据库是最关键的两个环节，利用ETL将异构数据源中的业务数据经过抽取、清洗和转换、加载到数据中心，为CRM的应用提供高质量的数据。本文以期货CRM项目为依托，详细阐述了期货CRM数据中心规划与设计，ORACLE RAC数据库规划，ETL设计与应用，对ETL的设计与应用研究是本文研究的重点。本文主要工作成果如下： 1) ETL相关理论的研究，包括数据仓库、ODS、ETL技术等。研究了构建数据中心相关的数据仓库、ODS和ETL技术，主要包括数据仓库体系结构，ODS体系结构，ETL中的数据抽取、转换、加载、元数据管理、数据质量保证和实时ETL等相关技术。 2)期货CRM系统数据中心规划与设计。根据期货CRM的系统需求，规划和设计了数据中心，对ETL的采集流程、数据抽取模式、数据清洗转换方式和并发处理方式进行了规划，对ORACLE RAC数据库系统、存储系统进行了规划和设计。 3) ETL在期货数据中心中的应用。根据期货数据中心的需求和特点以及ETL的相关技术，本文设计了一个具有并发和分布式处理能力、支持多种数据源、支持多种数据抽取方式和转换方式符合期货数据中心建设要求的ETL过程，并将该ETL过程应用到期货CRM数据中心，该ETL具有较好的扩展性，在支持多种异构数据源、并行和负载均衡、异常处理和质量保证等方面都具有一定的创新性。
[Abstract]:With the continuous development of the futures industry in recent years, the futures industry has become more and more dependent on IT technology. OTC systems, liquidation systems, asset management systems, office automation systems, and financial systems have all been built one after another. However, at present, the application is still mainly based on business operation, and the information among the systems is lack of effective means of sharing, forming a large number of "information isolated islands", lacking powerful business support, management control, business analysis and decision support functions. In order to achieve greater development, futures companies are actively changing their thinking, from external expansion to internal management, from "business-oriented" to "customer-centric." From "commission price competition" to "service quality competition", futures companies must attach importance to the core value of the company if they want to stand out in the fierce market competition. For futures companies, customer is the real value, customer relationship management (CRM) as a way to enhance customer satisfaction, improve customer loyalty, in-depth development of customer value effective application has been more and more attention of domestic futures companies. It is urgent to build CRM system based on data center. For CRM system, the most important thing is to enrich the accurate customer data. In order to meet the needs of CRM, to establish a unified customer view and to build an analytical enterprise data center is the foundation. In the process of data center construction, ETL extract transform and load (data extraction, conversion and loading) and database are the two most critical links. The business data in heterogeneous data source is extracted, cleaned and converted by ETL, and loaded into data center. Provide high quality data for CRM applications. Based on the future CRM project, this paper expounds the design and application of the future CRM data center planning and application in detail. The design and application of ETL is the focus of this paper. The main results of this paper are as follows: 1) the research of ETL theory, including data Warehouse (DW). The data warehouse and ETL technologies related to the construction of data center are studied, including data extraction, conversion, loading, metadata management, data quality assurance and real-time ETL. 2) Planning and design of data center of futures CRM system. According to the system demand of futures CRM, the data center is planned and designed, and the collection flow, data extraction mode, data cleaning conversion mode and concurrent processing mode of ETL are planned, and the ORACLE RAC database system is designed. The storage system is planned and designed. 3) the application of ETL in futures data center. According to the demand and characteristic of futures data center and the related technology of ETL, this paper designs a data source with concurrent and distributed processing ability. It supports a variety of data extraction methods and conversion methods to meet the requirements of the ETL process of futures data center construction, and applies the ETL process to the futures CRM data center. The ETL has good scalability and supports a variety of heterogeneous data sources. Parallelism and load balancing, exception handling and quality assurance are all innovative.
【学位授予单位】：浙江工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP315;TP308

【参考文献】