Fig. 7

Disruption comparison for \(\overline{k}\) = 4 and the three data. We developed an equation to find a benchmark for the level of information loss, or the anonymization impact. Named as Disruption. This diagram shows four different datasets; Adult, Seer cancer, Heart disease data and Kasandr. Each dataset was anonymized by Top-Down specialization, Bottom Up Generalization, and our MDSBA. It was found that our method was the lowest Disrupted