Improving Multivariate Microaggregation through Hamiltonian Paths and Optimal Univariate Microaggregation

The collection of personal data is exponentially growing and, as a result, individual privacy endangered accordingly. With the aim to lessen risks whilst maintaining high degrees utility, variety techniques have been proposed, being microaggregation very popular one. Microaggregation family perturbation methods, in which its principle aggregate records (i.e., microdata) groups so preserve through k-anonymity. multivariate problem known be NP-Hard; however, univariate version could optimally solved polynomial time using Hansen-Mukherjee (HM) algorithm. In this article, we propose heuristic solution inspired by Traveling Salesman Problem (TSP) and optimal solution. Given dataset, first, apply TSP-tour construction generate Hamiltonian path all dataset records. Next, use order provided given permutation records) input algorithm, virtually transforming it into solver call Multivariate (MHM). Our intuition that good solutions TSP would yield paths allowing algorithm find problem. We tested our method with well-known benchmark datasets. Moreover, show usefulness approach protecting location privacy, real-life trajectories datasets, too. compared results those best performing solutions, proposal reduces information loss resulting from microaggregation. Overall, suggest counterpart ordering microdata proper applying an leads reduction error keeping same guarantees.