| Along with the rapid development of the internet and various communication technology,information sharing is becoming more and more easily. Country, enterprise and individual cancollect the useful information more easily. At the same time with develop of the data miningtechnology and the application of the data releasing, people pay more attention toprivacy-preserving problem. In the past most studies were focused on privacy-preserving ofthe static data set, however, in the practical application, release data set is usually dynamic.How to protect privacy of dynamic data is key point of research.This paper study mainly the privacy-preserving problem of data set with internalattribute value updated. In order to protect the privacy of the types data sets, this paperintroduced the anonymous technology, barrel technology, and based on these technologiesputs forward λ-variety algorithm: Firstly, determine the type of attribute value update throughreading sensitive attribute field type from data table. Secondly, for different types of data setupdate, using different barrel create technology and record distribution method. Finally,divide equivalence class and publish. In addition, this paper considers the factor of dataaccuracy and puts forward the （D, λ）-variety algorithm which uses greedy thought, whichnot only protect privacy of release data set but also ensure the availability of the released dataset. The last part of the paper adopted Income data set and OCC data set from http://ipums.orgwebsite for experiment, and the results show that the proposed λ-variety algorithm and （D,λ）-variety algorithm can protect privacy of having external update and different categoryinternal attribute value update data sets. Compare the proposed λ-variety algorithm and theproposed（D, λ）-variety algorithm,（D, λ）-variety algorithm can ensure the accuracy of databut has lower privacy-preserving. So, using λ-variety or （D, λ）-variety algorithm can dependon the request of specific situation. |
PDF Full Text Request