【作者】 张德超；

【导师】 胡燕；

【作者基本信息】 武汉理工大学 ， 计算机应用技术， 2007， 硕士

【摘要】 长期以来，电信企业大量而详尽的电信业务数据只被简单的应用在各种业务系统中，而没有被更有效的开发利用。随着电信市场的竞争日趋激烈，如何利用这些数据为电信业提供决策支持服务，从而进一步拓宽电信业务，促进电信业务发展，成为当务之急。因此有必要建立一套基于全省综合业务的支持系统，该系统针对电信业务种类多，业务系统更新周期短的特点，利用数据仓库技术将大量电信数据转换成可靠的、商用的决策支持信息，并在数据仓库基础之上进行多维数据分析、数据挖掘和建模分析，帮助明确决策目标和进行问题的识别，了解经营情况，查看、细分客户群，从而帮助决策者科学地进行投资、开拓新业务产品等决策。本文主要研究数据仓库、数据挖掘等技术在电信行业经营分析中应用的相关理论和实际应用。本文首先对数据仓库，数据挖掘和DSS技术进行了系统的阐述，主要介绍了它们的概念和特点，体系结构，方法，工具等方面，为论文中讨论的议题进行知识准备；其次利用湖北电信立项实施企业级电信综合业务支持系统的机会，联合大唐公司和Longshine公司，通过对各部门级的数据进行整合，并结合多方面业务系统的数据，提出构建电信企业级的综合业务支撑系统模型的实现方法和技术要点来实现电信综合业务运营支持系统；再次通过对数据挖掘技术的经典算法—Apriori算法的研究，分析电信企业内部运营系统中已存在的海量历史数据，挖掘数据中的关联关系，从中提炼出业务发展情况和客户消费行为相关联的核心信息，找出潜在消费特征，消费群体等，为企业不同层次的管理者提供广泛的管理和决策支持，从而使其在激烈的市场竞争中谋取一席之地；最后，利用改进后的关联规则算法对湖北电信移动市话—小灵通的海量呼叫做试验分析，得出有效地改进结果—利用改进后的关联规则算法，能够有效地降低传统的关联规则算法中因支持度阈值和置信度阈值微小变化而引起的规则数量的巨大变化，使得对用户数量少得资费信息进行挖掘成为可能。更多还原

【Abstract】 For a long time, large and detailed data of telecommunications enterprises has only been applied in simple operational systems without being used more effective. With increasingly fierce competition in the telecommunication market, how to use these data to support decision and develop telecommunication business for the telecommunications industry become a focus. Therefore, it is necessary for the enterprises to establish a comprehensive business support system, which was aimed at the features of mulriple telecommunication business and the short update cycle. The system uses warehouse technology to convert larges numbers of telecommunication data into reliable and commercial information for decision, and based on data warehouse the system should be able to analyze the multidimensional data, data mining and modeling analysis. So it can help enterprise to decide its policy objectives, recognize problems, learn about its management situation understanding business and its customers. This will help decision makers to invest or develop new products and other business decisions reasoningly.This thesis mainly did some research of data warehouse and data mining technology in the telecommunications industry and its business analysis application of the theory and practical application. And it firstly systemic described those theory and technology about warehouse, data mining and DSS, which mainly introduced their concept and characteristic, system structure, method, tools, etc. The subject which discussed for the paper was carried on the knowledge preparation; Next, in this article, it united the DaTang Corporation and Longshine Corporation, carries on the conformity for various departments level data, and unified the various operational channel data, proposed constructs the telecommunication enterprise level comprehensive service support system model realization method and mainly technical points for the realization of the telecommunication synthetical service operation support system, in the opportunity of setting up the implementation enterprise level telecommunication synthetical service support system for Hubei Telecom; And thirdly, according to the research of the technology of data mining algorithms--apriori classical algorithm, it analysed those massive historical data of telecom operation and the correlation between data mining. Attained from the business development and consumer behavior related to the core client information, identify potential consumer groups, consumption characteristics for a wide range of enterprise managers at different levels of management and decision support, which could gain a foothold in the fierce market competition;Finaly, it did some analysis for those HuBei Telecom PHS with the improved data mining algorithm, and got an available improved result---which could reduce large change for the support and believe theshold value change’s reason in the traditional relationship algorithm. So it could make it possible for those few customers’ data mining.更多还原