PROJECT TITLE :

A Genetic Approach to Statistical Disclosure Control

ABSTRACT :

Statistical disclosure control is that the collective name for a vary of tools employed by information providers like government departments to shield the confidentiality of individuals or organizations. When the published tables contain magnitude information like turnover or health statistics, the popular method is to suppress the values of sure cells. Assigning a value to the information lost by suppressing any given cell creates the “cell suppression downside.” This consists of finding the minimum cost solution that meets the confidentiality constraints. Solving this downside simultaneously for all of the sensitive cells in a table is NP-hard and not possible for medium to large sized tables. In this paper, we describe the event of a heuristic tool for this problem which hybridizes linear programming (to unravel a relaxed version for a single sensitive cell) with a genetic algorithm (to hunt an order for considering the sensitive cells that minimizes the ultimate price). Considering a vary of real-world and representative “artificial” datasets, we show that the method is in a position to provide relatively low cost solutions for a lot larger tables than is attainable for the optimal approach to tackle. We show that our genetic approach is in a position to significantly improve on the initial solutions provided by existing heuristics for cell ordering, and outperforms local search. This approach is then extended and applied to massive statistical tables with over 20000zero cells.