Research And Implementation Of Data Extraction-Conversion-Loading (ETL) System In Medical Data Warehouse

In view of the massive data accumulated in various information systems used in the built hospitals,this thesis proposes how to use various information means to collect and mine the existing data to provide effective technical support for patients' treatment.Because of this,the collected data of medical expenses should be preprocessed and transferred to the corresponding data warehouse,then the online comprehensive analysis and summary should be carried out,and finally the effective data mining should be carried out.In this way,we can further analyze the potential relationships among patients,drugs,treatments and doctors.The discovery and use of such knowledge can form a powerful support and auxiliary role for more decision-making of the hospital,and also provide reliable technical support and valuable experience for further processing of other medical data in the future.This thesis mainly describes the necessity and urgency of promoting hospital informatization,and focuses on the problems of data sharing in the field of medical informatization.In view of these problems,this thesis proposes a general data model based on "five elements",and according to the characteristics of the medical data in the hospital data center,carries out the data Process research and design.Based on MapReduce ideas and methods,this thesis proposes a data mapping and collection storage method based on key value,and uses the general self-defined function designed by this thesis to complete the data collection of "empi".This thesis proposes the ontology based metadata management method,which can effectively deal with many problems in metadata semantic constraints and semantic information,and can provide more support to the medical field where metadata changes greatly.In the design process,this thesis uses the star and extended snowflake model to establish a data mart for various topics to meet the needs of decision-making analysis in various business areas.Through the above work,the requirements analysis,design,implementation and verification of the smart hospital system / platform based on the data center are basically completed,and the practice in our hospital has proved that the above methods and technologies are effective.