Outliers Eliminating for Enhancing Data Utility in Data Publishing

Provided by: Binary Information Press
Topic: Data Management
Format: PDF
In order to protect effectively the individual privacy, increase data utility and decrease information loss, anonymous processing of data is needed in privacy preserving data release and a k-anonymous data publishes model based on density clustering with outlier detection was proposed. The outlier existing in data set was fully considered in the clustering process. Privacy protection and data utility are achieved under the condition that total information loss of anonymous data is minimums. This model consists of as follows. First, the authors part data sets into clusters using density clustering while were eliminating outlier interference. Second, adjust clusters size and remove outlier again, which making the total information loss minimum.

Find By Topic