一种异构数据源模型转换和模式集成框架  

A Framework for Model Transformation and Schema Integration of Heterogeneous Data Source

作　　者：王博[1] 郭波[1] 

机构地区：[1]国防科学技术大学信息系统与管理学院,长沙410073

出　　处：《计算机科学》2007年第10期129-132,共4页Computer Science

摘　　要：异构数据源集成系统的主要任务是屏蔽数据源数据模型的异构性,提供对数据的统一访问。公共数据模型、模型变换和中间数据模型被广泛用来解决该问题。由于数据集成工作的工作量和任务对象日益复杂、规模日趋庞大,仅采用公共数据模型不能满足现有集成要求。异构数据源数据模型内在的关联性虽然使得面向语义集成变得更加复杂,但更利于模型转换方法在数据集成中的应用。模型转换是模式集成的基础,本文给出了异构数据源模型的形式化描述方法和基础模型转换操作形式化框架。该框架能够保证模型、实例和约束三者的相互独立,适用于大多数基础数据模型及其之间的转换和集成应用。通过该框架可简化和规范异构数据源数据模型转换和模式集成过程。The main task of heterogeneous data source integration is to providing a uniform way for accessing data source without considering the heterogeneous characteristic of data source. Common data model, model transformation and middle data model are widely used to solving this problem. As the work of data integration becomes more and more complicated, common data model can not satisfy the need of data integration now. Although works of semantic integration becomes more complicated for the inner relationship between data sources, but the transformation of model becomes much convenient, model transformation is the foundation of schema integration, in this paper we give a formal method for the description of heterogeneous data source and the formal framework for the primitive transformation between basic data model, this framework can maintain the independency among model, instance and constraints, which is more suitable for transformations between data models and the application of integration, the process of the transformation between heterogeneous data sources and schema integration will be simplified and formalize by using of this framework.

关 键 词：异构数据模型 模型转换 模式集成 

分 类 号：TP311.134[自动化与计算机技术—计算机软件与理论]