Differential privacy measurement method based on graph theory and mutual information amount

• Main Authors: HE WENZHU, WANG MAONI, PENG CHANGGEN, DING HONGFA, DING XING
• Format: Patent
• Language: Chinese; English
• Published: 15.11.2019
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention discloses a differential privacy measurement method based on a graph theory and a mutual information amount. A differential privacy protection framework is reconstructed by an information theory communication model, a differential privacy information communication model is constructed, an original data set is expressed as an information source, a published data set is expressed as aninformation sink, and a query mechanism and a noise mechanism are expressed as communication channels; the proposed differential privacy measurement model is based on an information communication model. A mutual informatization calculation method of privacy leakage amount is given by combining the characteristics of a graph with information entropy, the boundary of the privacy leakage amount onlydepends on the attribute number, the attribute value number and the differential privacy budget parameter of an original data set, and for the original data set which is randomly distributed, an opponent with any attack capab