Research On Information Retrieval And Privacy Preserving Methods For Medical Big Data

With the rapid development of medical informatization,China's smart medical system is formed initially,especially achieve remarkable results in the fields of health management.The establishment of a smart medical system marks the start of the era of medical big data.However,the utilization rate of massive medical data is currently low.On the one hand,the huge objects stored in the medical database system make the efficiency of retrieval low,which is the bottleneck problem of limiting the application of medical information.On the other hand,medical data has sensitive attributes,which makes it difficult to process data.If data were leaked,it will have a big impact on individuals and related institutions.Faced with massive medical data,it is a very important research hotspot how to efficiently obtain data and ensure privacy security in the use of health data.For solving the problem of the medical data retrieval efficiency and privacy protection in the background of medical big data,this thesis studies the information retrieval and privacy protection methods of medical big data,and proposes a new retrieval method and privacy protection mechanism in order to search quickly for massive medical big data and to prevent privacy breaches during the information retrieval process.The specific research contents are as follows:1.Aiming at the characteristics of various types of medical data and large amount,combined with R-tree index,a method for quickly retrieving medical data is proposed.This method uses the improved k-means clustering algorithm to construct R-tree clustering model through R-tree index.Search medical data to improve the retrieval efficiency of the system.2.In allusion to the problems of the high sensitivity of medical data,this thesis is proposed a privacy protection mechanism based on local differential privacy.Using the method,a small number of sensitive points are extracted from the collected raw data and added noise,and the data stream is reconstructed based on the disturbed sensitive points that is stored in the database.Information retrieval is carried out in the newly generated data set,and the results of the query statistics are sent to the users.The introduction of this mechanism can meet the privacy protection needs in medical data retrieval.3.On the basis of the smart medical information management system,experiment on the above two studies.It shows that compared with the performance of the system with hash index,the application of R-tree index can improve the efficiency of information retrieval.The protection mechanism used local differential privacy protects the privacy data,while the data is also available.