【作者】 胡璟超；

【导师】 唐建雄；

【作者基本信息】 武汉理工大学 ， 计算机应用技术， 2007， 硕士

【摘要】 近年来，我国通讯市场的拆分、重组使得通讯企业之间的竞争越来越激烈，企业的决策对通讯行业格局的影响显得越来越重要。通讯企业的经过近年的发展，逐步建立了比较完善的计费系统、结算系统、渠道系统、客服系统、网管系统、财务系统、统计系统等各种业务支撑系统，积累了庞大数量的各类数据。如何整合历史数据、构建商业智能(Business Intelligence，BI)平台，发挥对企业决策的支持作用，成为通讯行业普遍面临的一个重要课题。整个BI体系结构，从简单的报表输出和领导信息系统(Executive Information System，EIS)、多维分析统计和数据挖掘，到探索能力以及可定制的分析应用技术等，有了相当的发展。支持和维护任何BI环境的基础是建立一个高质量、可信赖的数据仓库。数据仓库技术的不断发展，给电信BI建设带来了前所未有的机遇和挑战。面对巨大的市场潜力和激烈的竞争，如何为电信企业构建一个先进、完善的数据仓库系统—即经营分析系统—进行决策支持，从而提升核心竞争力、进一步增强客户满意度已经成为国内电信企业发展的当务之急。本文在比较研究了典型的、流行的数据仓库架构的基础上，提出了适于大型企业的数据仓库集线器与轮辐式架构的改良方案—基于数据中间层的数据仓库体系架构，它可以增强数据仓库的可用性、易用性和可扩展性，实现各数据集市间的数据共享，降低ETL脚本开发的难度，以空间换时间的方式，提高系统的访问速度，减少系统资源的消耗。本文以省级电信公司经营分析系统为例，对其实施背景、业务目标进行了综述。重点论述了利用数据仓库技术、结合商业逻辑，构建电信数据仓库的中央数据仓库数据模型、中间层数据模型及数据集市层数据模型的设计思路、原则、约定和方法等。更多还原

【Abstract】 In recent years, the resolution and reorganization of China’s telecommunications market have prompted the telecommunications industry more competitive. Enterprise decisions have increasing influence on the telecommunications industry pattern. After years of development, Communications firms have been gradually building comparatively complete accounting system, settlement system, channel system, customer service system, network management system, finance system, statistic system, and other operational support systems, these systems have accumulated a huge amount of different types of information. How to integrate historical data and build BI ( Business Intelligence ) platform for the enterprises to play a supporting role in the decision-making, is a major task which the telecommunications industry are facing.The entire BI architecture, from simple report forms output system and the EIS (Executive Information System) , to multi-dimensional analysis statistic and data mining, to the ability to explore and customizable analytical application technology, has developed considerably. The starting point of supporting and maintaining any BI environment is creating a high-quality and reliable data warehouse. The development and maturation of data warehouse technology have brought unprecedented opportunities and challenges to BI application of the telecommunications. Faced with huge market potential and drastic competition, how to build a advanced, perfect data storage system (Operating System) for telecommunications enterprises to support decision, thereby enhancing their core competitiveness, enhancing customer satisfaction, has become a prior problem of the domestic telecommunications enterprises of the moment.Based on the comparative study of the typical architectures of the data warehouse, the modified project of centralized architecture which is suitable for the large enterprise data warehouse and is composed of central data warehouse layer, the middle layer, data mart layer, is promoted. The three layer data warehouse architecture project can enhance the reusability, flexibility and expandability of data warehouse, realize the data sharing of data marts , reduce the difficulty of ETL script, use space for time to improve system performance and to reduce consumption of system resource. Then the management and analysis system of a provincial level domestic telecommunications company is used as an example, the implementary background and business objectives were reviewed. On the basis of the use of data warehouse technology and commercial logic, The thesis discourses on the design ideas, principles, methods and agreement of the central data warehouse data model, the middle layer data model and data mart level data model of the telecom Data Warehouse.更多还原