Smart Health Service Platform Design And Its Key Technologies

The boom of IOT, mobile Internet and data mining etc, brings a dramatic change to health services. The form of health services become more and more informationized, personalized and intelligentized. But there still are many limitaions of the health services offered by different orgnizations. For instance, the data of different health services is isolated, so it can not be shared for each other. Another example, there is a big barrier of data flow through one health service to another, because the data format lacks of a unified definition. In regard of the problems above, this thesis designs a Smart platform of health service which is able to make different health services to work together in a smart and co-operative way.This thesis describes the network structure of the new designed platform and its data format and proposed the abstract definition of two kind of services the platform offered: real-time health service, delayed health service. This thesis introduces the basic theory of machine learning technology. Using the machine learning technology we realized a service, that giving the guide for people who has no idea of which department to go before they go to see a doctor on the new designed platform. This thesis introduces an advanced solution for distributed storage--HDFS as the data storage solution of the platform and gives the detail description for how to transplant the HDFS to our platform. This thesis introduces the defacto standard of distributed computation, the MapReduce programming model and gives the detail description for how to transplant MapReduce model to our platform. Adopting HDFS and MapReduce technology, the platform is able to deal with big data problem. Using distributed processing technology we realized a delayed health service which can extract the weighted vector data form the huge amout raw case report of illness.