期刊名称：Eastern-European Journal of Enterprise Technologies
印刷版ISSN：1729-3774
电子版ISSN：1729-4061
出版年度：2017
卷号：5
期号：4
页码：31-39
DOI：10.15587/1729-4061.2017.113046
语种：English
出版社：PC Technology Center
摘要：In the work, a modification of the method for solving the task of providing data group anonymity is proposed, which implies automated solution selection without expert participation. Modification lies in identifying solutions to the task, in which outliers are detected automatically and don’t match the outliers in the initial distribution of the information about the group of respondents. Thus, automating the solution selection improves data group anonymization efficiency by reducing the time necessary for their analysis for masking sensitive features of the distribution.Testing the developed modification is done by solving the task of masking regional distribution of military personnel in the state of New York. As a result of solving the corresponding group anonymization task, 1,000 solutions were obtained. It is established that only 24 out of 1,000 solutions, or 2.4 % of the total number, are feasible, i. e. the ones in which all the outliers are masked. Automated selection of such a small number of solutions is significantly faster than the manual approach, which speaks in favor of the proposed modification for improving data group anonymization efficiency.
关键词：memetic algorithm;group anonymity;microfile;outlier;modified Thompson tau technique
