基于CWM的结构化异构数据抽取方法研究与实现  

STUDY AND IMPLEMENTATION OF STRUCTURED HETEROGENEOUS DATA EXTRACTION METHOD BASED ON CWM

作　　者：夏秀峰[1] 孙娜[1] 石祥滨[1] 徐蕾[1] 

机构地区：[1]沈阳航空工业学院计算机学院,辽宁沈阳110136

出　　处：《计算机应用与软件》2009年第12期108-110,120,共4页Computer Applications and Software

基　　金：辽宁省自然科学基金(20052007)

摘　　要：传统的信息集成虚拟方法通常采用Wrapper技术实现结构化异构数据的抽取。针对Wrapper与结构化异构数据源间一对一关系的非灵活性,提出了一种基于公共仓库元模型CWM(Common Warehouse Metamodel)的信息集成架构,利用CWM独立于任何具体实现的优势,在其基础上生成的Wrapper程序可以很好地适用于与之相关的所有结构化异构数据源的抽取。CWM元模型所提供的统一文档类型定义(DTD),使结构化异构数据源拥有了统一的XML表达方式,从而使抽取后的信息可以进行更有效的共享和转换。Traditional virtual methods of information integration often use Wrapper technique to implement structured heterogeneous data extraction. In light of inflexibility of one-to-one correspondences between Wrapper and structured heterogeneous data source, an information integration architecture based on CWM is proposed. Taking the advantage of CWM in independent to any specific implementation, the Wrapper program generated based on CWM can apply well to all structured heterogeneous data sources extraction related to the program. Uniform document type definition （DTD） offered by CWM metamodel makes structured heterogeneous data sources hold a universal XML expression, so the extracted information can be shared and transformed more effectively.

关 键 词：结构化异构数据 信息集成 虚拟方法 公共仓库元模型 XML 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论] TP391.4[自动化与计算机技术—计算机科学与技术]