Research And Implementation Of Medical Information Privacy Protection Technology Based On K-Anonymity

Purpose: With the strengthening of medical information construction and the need for scientific research,a large amount of medical information is being collected and shared.There is a lot of potentially valuable information in medical information.It can be beneficial and make an important contribution to human health and even the entire society if it be researched and used.If medical information is released without being processed or handled improperly,it may be divulging patients privacy and causing unnecessary trouble to them while the information used by people with ulterior motives.The issue of privacy protection has become particularly prominent during data sharing.Therefore,we must ensure two factors before the release of medical information.First,ensure that the private information in the released data will not be leaked.And second,ensure the validity of the published data for scientific research.Method: At present,privacy protection technologies mainly include data encryption technology,data distortion technology,and anonymity protection technology.Each of these three technologies has its advantages and disadvantages.Data distortion technology has high efficiency but loss too much information.Data encryption technology has good security but slow operation speed and high operating cost.Anonymity protection technology is relatively balanced in all aspects.It is not only free from large calculations,but also does not cause too much data loss.Anonymity protection technology is an important technical means in the field of privacy protection.The idea is to change some of the data in the original data table by means of generalization,suppression,etc.,so that the processed data table cannot be linked with other tables to infer personal information and avoid link attacks.Therefore,this article chooses to study the anonymization technology.K-anonymity model is a classic model in the anonymization technology,and the Datafly algorithm is the classic algorithm of k-anonymity model.So I improve on the disadvantages of the Datafly algorithm and proposes a multi-attribute generalization algorithm.Achievements and conclusions: Aiming at the problem of selecting single generalized attribute in the Datafly algorithm and the situation that the most value attribute may not unique,we propose a multi-attribute generalization algorithm.Compared with the original Datafly algorithm,it improves the data accuracy after k-anonymity processing.Therefore,it has more application significance and is suitable for privacy protection of medical information forms.