【作者】 潘建军；

【导师】 王加阳；

【作者基本信息】 中南大学 ， 计算机科学与技术， 2011， 硕士

【摘要】 伴随着人类社会信息化的发展,信息呈现出爆炸性增长的趋势。企业业务数据作为信息的关键元素分布情况也呈现出越来越复杂的现象,其中主要包括企业内部数据冗余、业务部门之间数据共享困难等。数据集成技术的出现主要是为解决信息系统内部异构数据之间相互访问困难和统一管理困难的问题。在研究数据集成技术的过程中,本文参考了传统数据集成系统的设计模式,并且分析各种已存的数据集成技术的优劣关系,依据技术发展趋势和业内对数据集成系统提出的新的要求,将研究重点转向了SOA构架技术在数据集成中的应用。通过分析比较各种实现SOA构架的技术,本文选用具有良好的互操作性、封装性、松耦合性和具有行业规范标准的Web Service技术作为实现该数据集成系统的主要技术手段。该数据集成系统中信息交换以ESB系统为依托,参与完成数据集成过程的各种功能组件,例如智能消息路由模块、语义冲突消解模块、查询处理模块等,都以系统服务组件的形式集成于该数据集成系统之中,所有的功能组件都具有即时加载即时生效的能力。在现实应用条件下,由于不同数据服务提供商可能为相同的功能提供不同版本的服务模块,而服务请求者在对数据服务发出请求的过程中也可能具有不同的侧重点,因此本文在服务的选择问题中设计了一种基于服务质量模型的智能路由模块,用于解决消息传播过程中的路由选择问题。另外,本文还重点讨论了数据集成系统的安全模型和语义冲突消解模块的设计原理和设计方法。本文通过一系列的理论研究及实践工作设计并开发了一套能为用户提供安全、健壮、可靠并且具有较强应变能力的离散数据集成系统。更多还原

【Abstract】 Along with the development of social informatization, the information shows a trend of explosive growth. As a key element of information, enterprise business data also has emerged more complex phenomenas such as enterprise internal data redundancy and hard data sharing in different business departments. The appearance of heterogeneous discrete data integration technology is primarily to resolve the problems of mutual visitation and unified management in different information systems. In the process of researching data integration technology, the research not only refers to the design patterns of traditional data integration systems but also analyzes the advantages and weaknesses of the existing technologies. According to the technology developing trend and the new demands on data integration system, this thesis focuses on the service-oriented architecture technology (SOA).By analyzing the different realization technologies of service-oriented architecture, the research selects web service as the main measure for achieving the data integration system, because web service has good interoperability, encapsulation, coupling and industry specification standard. The data integration system exchange messages based on enterprise service bus. All of the functional components of data integration, such as intelligent message routing module, semantic conflict resolution module, query processing module are integrated in the system as components. These components of the system have ability to load and take effect in time.In fact, different data service providers may provide different service modules for the same function. Service requestors may have different needs for themselves during the data service request. Therefore, In order to solve the routing problem of message spread and expansion, an intelligent routing module based on service quality model is designed. In addition, the research mainly discusses the design principals and methods of the security module and semantic conflict resolution module of data integration system. Through a series of research and practical works, the thesis designs a data integration system with good safety performance, reliability and adaptability.更多还原