Data Extraction Conversion Tools, Data Mapping Design And Its Key Technologies

Nowadays, with the trend of data concentration, service comprehension, and scientific decision in information construction, most of the heterogeneous legacy systems are incompetent to meet these requirements. More and more companies and enterprises require data integration and data exchange. In order to deal with the complex data processing problem in data integration and data exchange, a specific tool for data extracting, transforming and loading (normally called ETL tool) has shown up. ETL tool has played an important role in data integration.The paper first introduces the background knowledge about ETL tools, and then analyzes the present research status of them. With the conclusion that current researches haven't done much work on mapping relationship between source and target data, despite of its core status in data transformation process, chapter 1 brings forth the research goal and research content on it.According to the research goal, the paper designs a corresponding ETL tool, ETLA (short for Extract -Transform-Load-Analysis). In ETLA, user defines ETL task by establishing data mapping relationship between source and target, then the system will convert the mapping relationship to execution script. Chapter 2 exposes the system framework, functional partition, data process method and data mapping procedure.In addition, this paper emphasizes the data mapping relationship in chapter 3. After having analyzed various data mapping relationships, the paper provides a formalized description for Data Mapping. In ETLA, mapping expression signifies the data mapping relationship. The mapping expression composes three parts: Source Data Elements, Target Data Elements, and Relationship among Source Data Elements, the express rules of the three parts being implemented. After data mapping relationships are established, the ETLA data transformation module will convert the mapping relationships to execution scripts, and execute it.The last part of the paper designs a mapping-index repository. The repository organizes source data index by topics. With the help of therepository, the user can locate the source data as fast as possible, and establish the data mapping relationship between target and source.