摘要
为了将频繁产生的分布在世界各地的金融数据按需地高性能集成,提出了基于ETL(extract-transform-load)的金融数据集成过程模型。对规则引擎原理进行了研究,建立了基于业务转换规则的插件式扩展实现可复用可定制的业务转换过程。利用事件驱动的交互模型和元数据映射保证非结构化和半结构化数据之间无差异集成,采用增量式数据处理解决数据集成中棘手的性能问题。通过实践项目的验证,对比传统数据集成方法和该过程模型,验证了该过程模型的有效性。
In order to integrate the distributed data frequently with high performance, an ETL （extract-transform-load） based data integration model is proposed. Business rule has been studied and integrate into this model to get business rule as plug-in transform solution. On demand request and Meta data mapping can shield the difference between the structured data and un-structured data. Performance issue has been analyzed and the incremental integration is introduced to extract data source in an incremental way. Based on practical project, traditional point to point integration is compared to this model to illustrate this model can have high performance and easy to be implemented.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第9期2070-2072,2104,共4页
Computer Engineering and Design
