Design And Implementation Of Medical Information Publishing System Based On Full-Domain K-anonymous

Produce large amounts of data resources during the process of medical diagnosis,need to be reasonable induction,analysis and mining.However,there is a lot of patients' privacy information,physically or psychologically.Disclosure of this information may do great harm to the owners.Therefore,privacy safety during medical data releasing has attracted wide attention.Currently,it's not appropriate to apply the common three privacy-protecting methods,namely property deleting,data encryption and data perturbation(randomization),to medical data release.The main reason is that not only the availability of data after processing,but also the avoiding of user's privacy disclosure should be ensured during the releasing.It's hard to do both with the three mentioned methods.Based on the above issues,K-anonymity has been proposed,which is through generalization or hiding the original data to form the capacity of at least K clusters,in order to make each tuple not be distinguished from at least K-1 individuals.The algorithm can reduce information loss while projecting patients' privacy.However,it remains to be improved when it comes to classic algorithms to implement K-anonymity.Among global anonymity algorithm,Incognito is a kind of classic exact solution algorithm.It obtains the minimum amount of information loss globally by traversing the lattice distance vector generalization optimal solution,which can guarantee data availability while protecting privacy after medical data release.But with low time efficiency,it is not applicable in processing large-scale data anonymously.Meanwhile,with the increasing of quasi-dientifier property,the time to be solved will be increased dramatically.According to time efficiency problem,first of all,we combine identity law of data set with Incognito algorithm,to propose F-Incognito algorithm,a new global anonymity algorithm.Secondly,set up the experimental environment,proved by simulation experiment,the algorithm was able to retain Incognito algorithm characteristics,while improving the generalization of solving node table time efficiency,shorten time to as high as 60%.Thus,F-Incognito algorithm when dealing with large data sets,has a significant advantage.The last,base on F-Incognito algorithm,we design an anonymous medical data release system.