T He D E S Ig N A N D I Mp L E Me Ntati O N O F D Atae X T R A C T I On S U B Sy St E M Of D A Ta S Ha R I N G Pl Atfo Rm

With the popularization of computer skills and the rapid development of information technology, the source and talents as academic cradle the institutions of higher learning, the use of digital information technology for university teachers and students provide quicker and more high-quality information services, gradually achieve the digital campus form is the universities into modern the marks of information times, while "data island" phenomenon is the current domestic universities the inevitable outcome of the digital campus construction, in order to solve the data island from the message of the barriers, meet the new needs campus informatization, must establish a unified data sharing platform as the data layer and the application layer bridge between the reunification of data standards, to ensure that the data of authority, consistency, timeliness, for scientific research, teaching, management, providing the user data resources, provides comprehensive and integrated multi-sector global business services.Data extraction subsystems is the basis for data sharing platform, in order to meet the demand for top-level data sharing platform, data extraction subsystems to achieve the following:First, according to ETL (Extraction-Transformation-Loading, data Extraction, conversion and Loading) internal data transfer principle for data extracting work to establish a system for the model, the development work of after laid a unified model base, also can effectively the function of the software upgrade and iterative development maintenance.Secondly, establish a good data source control measures can be customized to monitor user-defined rules for monitoring objects, minimum size to achieve a field in a database table, and can set the duty cycle of extraction tasks, to achieve a central database of "active Pull "type of data extraction.Third, the actual needs of the environment according to the data set the appropriate mechanism for incremental data extraction, along with a snapshot of differential detection and log analysis and detection of two incremental data detection means, and strive to fit the scene in the appropriate method to use in order to bring more Efficient and more efficient service.Fourth, in the information services layer using flexible reporting tools JasperReort plug function allows users without programming, as long as intuitive and simple steps they can easily report the development and management.Fifth, to adapt to multiple users simultaneously, using spring transaction management technology to simultaneously view the same transaction, to prevent the emergence of conflict. And through the transaction management, to achieve the same user on multiple views simultaneously, and save the modified function temporarilyBased on software engineering, from system requirements analysis, system model design, programming and testing, we proposed a reasonable solution for these issues, and details of data extraction and realization of subsvstems.