海量医疗数据挖掘平台的研究与设计-通信与信息系统专业论文.docxVIP

武汉理 武汉理工大学硕士学位论文 万方数据 万方数据 Abstract With the rapid development of medical information, the traditional digital information systems and electronic health records systems of hospitals can not meet the massive medical data storage and medical data services requirements. After massive medical data generation and collection, how to efficiently store heterogeneous, massive, real-time, a variety of data in order to achieve fast and accurate response to large-scale complex health data queries; while applying data mining and knowledge discovery theory of medical history data modeling and analysis, how to mine key physiological characteristics from medical data, and reliable, fast and efficiently detect early disease and predict health risks for users to provide valuable medical services, which need to be solved in the current health areas. Based on the characteristics of mass medical data, and the current lack of existing massive medical data mining applications, the massive medical data mining design platform is proposed, which includes massive medical data mining platform model, rapid statistical query algorithm of medical data, as well as medical relevance of data mining. In this paper, the specific contents are as follows: Based on the characteristics of massive medical data and medical service requirements, massive medical data mining platform model is proposed, including design principles of data mining platform, massive medical data storage strategy, as well as massive medical data processing. The platform model has the flexibility, scalability, reusability and so on. Based on fast query response requirements of service built by data statistics of massive medical data and strong association characteristics of medical data, a fast statistical query algorithm of massive medical data based on statistical tree and incremental computing is proposed, which optimizes the storage model, improve query efficiency, and efficiently support data statistical based medical service,