铁路客票数据仓库建设方案的研究

Study on the Building Scheme of Railway Ticket Data Bank

该论文运用近代发展起来的商业智能理论 ,构造了铁路客票数据仓库建设的完整解决方案。从我国铁路客运营销决策支持需求出发 ,论述了建设铁路客票数据仓库的必要性 ,提出了客票数据仓库的逻辑结构以及硬软件环境方案和实施方案。围绕数据源组织、数据存储和数据展现等问题 ,对系统建设中的若干关键技术进行了探讨。论文提出了铁路客运数据仓库建设的框架结构 ,主要是利用客票系统数据作为主要数据源 ,通过操作数据库实现数据暂存和预处理 ,经过数据转换将数据加载到数据仓库中 ,并建立面向应用的数据集市 ,最终通过多维引擎和WEB引擎实现联机分析和决策支持。客运数据仓库建设的基础是完整的数据 ,为此要创建客运企业基础数据 ,实现分布式数据库环境下客票发售数据的广域网采集 ,并补充必要的外部数据 ;数据仓库建设的核心工作就是将客运管理和决策支持所需要的信息从日常客票交易产生的数据中分离出来 ,进行清洗、整理、综合、概括 ,按照便于访问和分析的新的数据结构进行存储。为此需建立维表和事实表 ,并通过维度的键值来实现维表和事实表的关联 ,利用星型结构为数据仓库建立完善的模型。数据仓库建设的目的是实现数据展现与决策支持 ,为此应借助业务智能软件工具访问数据仓库 。

Using the recent period commercial intelligent theory, this paper constructs a whole solving scheme of railway ticket data bank. Proceeding from the need to support the decision of railway passenger traffic marketing, the paper discussed the necessity to establish of railway ticket, and put forward the logical structure and soft hardware circumstance scheme, and executive plan. This paper also studies the some critical technology in system building, around the data origin organization, data storage and data exhibition.This article offers the frame structure of the building railway ticket data bank, utilizing mainly the data origin of passenger ticket system. Through the operation database, we realize the data temporary storage and pre-processing. Data is added to the data bank after conversation, and a data market facing, application is established and the online analysis and decision support is realized. The data bank is the whole data origin. So, the base data of passenger traffic enterprise, the wide area network acquisition of passenger ticket under the circumstance of distributed database is produced, and the necessary outer data supplied;the critical word of data bank is to depart the information of passenger traffic administration and decision needed from the data of passenger ticket sale, and store as new data structure style. So, it is necessary to build the dimension table and factual table. Through the key assignments of dimensionality, we can realize the conjunction of the two kinds of tables. Using the star structure, we build the perfect model of data bank. The goal of data bank is to realize data exhibition and decision support. Using the intelligent software, we can access the data bank and fulfill data analysis at different levels.