Study And Implementation Of ETL Technology For Complex Scientific Text Data

Scientific data is to engage in a variety of scientist on the basis of their scientific data in endless discovery of new laws, the discovery of new knowledge. Scientific data on the management of scientific research is an important guarantee for the work, good management can accelerate the progress of scientific research. However, the handling of scientific data is vast and complicated, so the academic community has to address the urgent need for new methods for managing scientific data.The scientific data storage is an important aspect of scientific data management. In the initial acquisition stage, as a result of limitation of equipment, the environment, performance and other reasons, the scientific data are saved as a text format, with particular advantages of semi-structured format that makes the storage more fast and more convenience. In scientific data processing and analysis stage, in order to avoid the disadvantages of the traditional version of the preservation of scientific data like ambiguity of data meaning, difficulties of management, high price of query and so on, the relational database technology is used to manage scientific data. provide a excellent tool for a variety of applications. In this case, the difference of the storage method between the two phases raise a question about how to transform the data format effective.At first, this paper analyse the structural characteristics of scientific data, with which this paper give model and a formal expression method for the structure and the extraction method based on that model. Then, for the characteristicses of text data and relational data, this paper develops a mapping model and method that connect the two format. It is very important of the establishment of the connection of two types of data format data model mapping methods and modelst.In this paper, a system is designed and implemented based on the contents discussed above, with the overall structure of the system architecture in accordance with the design of ETL structure with its features. Then, combining the characteristics of marine scientific data, the ETL technology is put into marine scientific data, and the processes of modeling and extracting are implemented. Then a mapping model is established for marine scientific data, and put the developed system into marine scientific data.