面向财务决策系统的数据建模和抽取研究

发布时间：2018-03-22 12:28

本文选题：数据仓库　切入点：维度建模　出处：《首都师范大学》2013年硕士论文　论文类型：学位论文

【摘要】：财务管理作为国内首先实现计算管理的领域,积累了大量数据,是管理的财富,对这些数据的分析有重要的意义。传统的基于数据库技术的决策支持系统由于其技术弱且缺少统一性,无法处理大量的数据且应用复杂,已无法满足用户的需要。数据仓库技术和联机分析处理(OLAP)技术的出现为大数据的处理带来了希望,也解决了传统的决策支持系统的问题。数据仓库技术为决策支持系统的数据存储和管理提供了解决方法,但数据的原始记录一般存储在不同的数据库和源文件中,这也使得原始数据直接导入到数据仓库是不切实际的。因此,首先我们要为数据仓库的构建建立模型,然后在设置的模型的基础上设计抽取工具,通过数据抽取、转换和加载(ETL)这一过程将源数据库中的数据装载到目标数据库中。ETL是构建数据仓库的重要组成部分。ETL这一过程需要解决如主题的分散、数据重复、同一对象的命名方式或表达方式不一致的现象等问题。随着计算机技术的快速发展,数据仓库技术也得以快速发展以及广泛的应用。本文基于某机关单位的总账信息管理系统和非税信息管理系统,详述了数据仓库的维度建模和ETL工具的设计。本文采用维度建模作为其建模策略,结合总账和非税管理项目,主要探讨了维度建模设计问题,其中包括事实表和维度表的设计,通过星型模型来进行维度建模。首先介绍了维度建模的理论基础,结合具体的项目来进行维度模型的设计。 ETL是构建数据仓库这一过程最重要的设计。首先介绍了ETL的结构及这一过程元数据的重要性。然后具体结合总账抽取工具和报表抽取工具,详述了基于元数据的ETL过程的设计。
[Abstract]:Financial management, as the first field to realize calculation management in China, has accumulated a lot of data and is the wealth of management. The analysis of these data is of great significance. Because of its weak technology and lack of unity, traditional decision support system based on database technology is unable to deal with a large number of data and its application is complex. The emergence of data warehouse technology and on-line analytical processing (OLAP) technology brings hope to big data's processing and solves the problem of traditional decision support system. Data warehouse technology provides a solution for data storage and management in decision support systems, but the original records of data are generally stored in different databases and source files. This also makes it impractical to import raw data directly into the data warehouse. So, first of all, we have to model the construction of the data warehouse, and then we design the extraction tool based on the set model. The process of loading data from the source database into the target database. ETL is an important component of building a data warehouse. Problems such as inconsistent naming or expression of the same object. With the rapid development of computer technology, data warehouse technology has been rapidly developed and widely used. This paper is based on the general ledger information management system and non-tax information management system. The dimension modeling of data warehouse and the design of ETL tool are described in detail. In this paper, dimension modeling is used as its modeling strategy, and combined with general ledger and non-tax management project, dimension modeling design is mainly discussed, including fact table and dimension table design. Firstly, the theoretical basis of dimension modeling is introduced, and the design of dimension model is carried out with specific items. ETL is the most important design in the process of building data warehouse. Firstly, the structure of ETL and the importance of metadata are introduced. Then, the design of ETL process based on metadata is described in detail by combining general ledger extraction tool and report extraction tool.
【学位授予单位】：首都师范大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP311.13;F233

【参考文献】