基于数据源自动侦测更新的并行视图维护模型  

A Parallel View Maintenance Model Based on the Automatic Detection of Data Source Updates

作　　者：左亚尧[1] 潘久辉[1] 舒忠梅[2] 

机构地区：[1]广东工业大学计算机学院,广东广州510631 [2]中山大学,广东中山510275

出　　处：《计算机工程与科学》2003年第4期50-53,共4页Computer Engineering & Science

基　　金：广东省自然科学基金资助项目(990460)

摘　　要：数据仓库体系结构是数据仓库建设和维护的重要理论基石,传统的体系框架简单易行,但不够完善。斯坦福大学提出的WHIPS模型解决了信息源自动侦测更新的问题,但由于模型自身存在的瓶颈,使并行更新处理产生阻塞现象。为此,本文提出了一个改进方案,引入了时间戳单元,增加了其中两个重要模块的并行处理能力,并给出一个修正的数据仓库系统结构。The architecture of a data warehouse is the theoretical basis for designing and maintaining the data warehouse. The traditional architecture is simple and easy to achieve,but it is not prefect.The WHIPS model presented by Stanford has solved the problem of automatic detection of data updates at different information sources. However,because this model has a bottleneck, a series of concurrent data updates occur,and the updates will lead to blocking.Hence this paper gives an improved solution which introduces timestamp units and increases the parallel processing ability of two key models in the WHIPS system.Finally, a modified architecture is presented.

关 键 词：数据仓库 数据库 体系结构 数据源 信息源 并行视图维护模型 自动侦测更新 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]