Design And Implementation Of An Assessment System For Privacy Disclosure Risk Of Medical Data Release

The release of medical data plays a significant role in data mining and data research and it promotes the development of smart medical care.Nevertheless,it also brings the risk of medical data privacy leakage.In the current data publishing scenario,there is still a problem that the data receiver uses the existing data or background knowledge to attack the released data,resulting in the leakage of personal privacy information.However,there is a lack of better methods and systems to assess the risk of medical data privacy leakage.Therefore,considering the above problems,this study proposes an assessment scheme for privacy disclosure risk and a privacy protection scheme for data release.Similarly,it also designs and implements an assessment system for privacy disclosure risk of medical data release based on the relevant theories and methods such as privacy preference,information entropy,and similarity distance.The specific research work is as follows:Firstly,design an assessment scheme for privacy disclosure risk of medical data release.Because of the lack of assessment methods for privacy leakage risk in data release,use Pearson Similarity to calculate the user’s subjective attribute sensitivity based on the user privacy preference matrix,and then obtain the objective attribute sensitivity by averaging it.Next,by constructing the distance formula,the original data and evaluation data are used for distance measurement to calculate the similarity of attributes.Information entropy and mutual information are used to calculate the correlation coefficient between attributes.And form the correlation of attributes by using the maximum value of the correlation coefficient with other attributes.Finally,the attribute privacy score is calculated by integrating the sensitivity,similarity,and relevance of attributes.Then construct the privacy risk index according to the score,and divide the risk level.The effectiveness of the assessment scheme for privacy disclosure risk is verified by experimental analysis.Secondly,design a privacy protection scheme of medical data release.Given the high risk of privacy leakage in the evaluation,the attribute privacy risk level is firstly divided by the attribute privacy risk score.Accordingly,the appropriate privacy protection method and protection strength are selected based on the attribute privacy risk level and the characteristics of the medical data to realize the adaptive privacy protection of the data.The scheme solves the problem of balanced privacy protection of discrete data and continuous data in structured medical data,and realizes the functions of anonymous privacy protection and differential privacy protection of data.Thus,the balance of data privacy protection can be completed and the effectiveness of data can be improved.Finally,design and implement an assessment system for privacy disclosure risk of medical data release.According to the application requirements for the assessment of privacy disclosure risk,based on the above solutions,firstly analyze the functional and nonfunctional requirements of the system and design the system architecture and technical implementation plan in a comprehensive manner.Then,from the user’s point of view,the main module functions are analyzed and designed in detail.And the above design schemes are applied to the privacy risk assessment module and privacy protection module respectively.Finally,complete the development and implementation of the system and test it.The system mainly includes a privacy risk assessment module,a privacy protection module,a data release module,and a user management module.The assessment of the risk of privacy disclosure of structured medical data is realized,which has a good effect on reducing the risk of privacy disclosure of medical data and realizing the sharing of medical data.