【作者】 文小敏；

【导师】 袁清珂；

【作者基本信息】 广东工业大学 ， 机械设计及理论， 2005， 硕士

【摘要】 ERP在企业中已获得了广泛的应用,然而ERP运行中积累的大量数据并没有得到充分的利用,挖掘ERP大量数据中蕴涵的知识是目前ERP研究的热点,也是ERP系统进一步深化应用的必然趋势。数据挖掘是从大量数据中挖掘出有规律的、揭示其内在必然联系的技术,是利用ERP大量数据中蕴涵的知识进行企业运营辅助决策的重要手段。 论文在分析了ERP和数据挖掘研究与应用现状的基础上,探索了ERP原理、数据挖掘技术和数据仓库技术的综合应用。 论文在数据挖掘一般理论的基础上,研究了面向ERP数据挖掘的基本原理,提出了面向ERP数据挖掘的体系结构。研究了面向ERP数据挖掘的算法模型,重点探索了关联规则算法、聚类分析算法和神经网络算法。论文将上述理论研究应用到ERP中销售数据挖掘过程,提出了ERP销售数据仓库和ERP销售数据挖掘的实现方法,阐述了销售数据挖掘的需求分析方法,分析了销售数据仓库的系统框架、数据挖掘结果的评估方法,重点讨论了神经网络预测销售的时间序列方法,建立了神经网络时间序列预测模型。分析了销售数据仓库的开发过程和方法,完成了数据仓库的概念模型、逻辑模型和物理模型的设计工作。在SQL Server 2000环境下,用SQL语言和Visual C++ 2005开发ERP销售数据挖掘原型系统。结合某企业的实际情况,将原型系统进行了实例应用,获得了较好的效果,验证了本文理论研究和实现方法的可行性。更多还原

【Abstract】 Although ERP has been widely used in enterprise, mass data accumulated during the running of ERP has not been taken full use of. Nowadays, mining the knowledgy implied in the mass data of ERP is a research hotspot and an inexorable trend of the application of ERP. Data mining is the technology that mines regular knowledgy out of massive data and reveals inherent, certain relationship. And it is also the important means by which enterpises take use of the knowledge implied in the data of ERP to assistant decisions in the operations of enterprises.The synthetic appplication of principles of ERP, data mining technology and data warehouse technology was discussed in this paper on the basis of analysing the current situation of ERP and data mining.The essential principles of ERP-Oriented data mining were researched, and the architecture of ERP-Oriented data mining was suggested in this paper on the basis of general theories of data mining. The algorithm model of ERP-Oriented data mining was researched and association rules algorithm, clustering analysis algorithm, neural network algorithm was given. All above-mentioned theories were applied into the process of mining sale data in ERP, and introduced the implementation approach of ERP sale data warehouse, ERP sale data mining and the method of requirements analysis of sale data mining was put forward, After analysing the system framework of sale data warehouse and evaluation method, gave the method of forecasting sale time series using neural network, established time series forecasting model using neural network.更多还原