Windows dna - m put forward by microsoft provides an integrated platform for developing next generation manufacturing applications ; opc provides series of standard interfaces to integrate field devices , industrial control and production management software seamlessly ; xml provides the same data schema to exchange enterprise information inside and outside 微软提出的windowsdna - m体系为开发新一代制造应用软件提供了集成平台; opc技术为现场设备、工业控制、生产管理软件间的无缝集成提供了一套标准接口; xml技术为企业内外信息交换提供了统一的数据模式。
A hdsisbs ( heterogeneous data sources integration system based on soap ) prototype is put forward to overcome above difficulties . it is shown that hdsisbs simplifies the conversion of various data schema to the data schema in integration layer and provides the pnp ( plug and play ) feature of data source . our thesis is organized as follows 针对以上情况,设计了一个基于soap的异构数据源集成系统原型,即hdsisbs ( heterogeneousdatasourcesintegrationsystembasedonsoap )原型,简化了各数据源数据模式与集成层数据模式的转换工作,实现了数据源的“即插即用” 。
This paper describes the basic features and components of data warehouse system , and deals how to use description - driven technology to integrate different data warehouse systems , how to implement the change from one data schema to another , how to clean dirty data in data transformation process , and how to exchange data among different components or systems . at last , this paper takes two products to illustrate how to implement systems following these principles and methods . these two products are the e - chain system as an application in commerce domain , and the ftedws system as an application in engineer test domain 本文分析了数据仓库系统软件的基本特征,提出了利用描述驱动技术来实现数据仓库系统的集成管理,描述了etl操作和分析处理的基本处理流程和相应的执行构件,定义了集成框架中数据模式转换规则和数掘清洗规则,构建了一个基于星型模式和对象模型的分析模型和相应的数据查询语言,提出了集成框架系统构件间的数据交换标准,并定义了基于此标准的的数据交换和元数据交换方法,探讨了集成框架标准构件管理的基本方法和权限管理,最后介绍了数据仓库集成框架系统在商业领域的应用实例e - chain系统和工程试验领域的应用实例ftedws系统。
The key to realize the organized integration of various data sources is how to specify all kinds of data by a unique data schema and mask the heterogeneity of their platforms and data structure , etc . there are some shortage in existing data integration systems : complex conversion of any data schema to the common schema ; little integration about semistructured data ; lackness of interoperability in bottom level communication mechanism based on software component 实现数据无缝集成的关键和困难是如何以一种统一的数据模式描述各数据源中的数据,屏蔽它们的平台、数据结构等异构性。已有的数据集成技术,存在一些不足:公共模式同各数据源模式的转换工作繁杂;对结构化数据的支持较好,而半结构化数据不够重视,数据源参与集成的程度不高;基于软件组件的底层通信机制缺乏互操作性。