Abstract:

One of the most widely used and most dynamically evolving model which intends to tackle statistical disclosure is differential privacy. The method described in this paper, based on the differential privacy principles, pro-poses adaptive metrics that together state an effective mechanism of pre-venting privacy violations by inference attacks. The method remains effective for the dynamic the information-based systems in which data volume increases and characteristics fluctuates significantly overtime and can be applied for the newly developed systems as well as already well-established systems as an additional software component.