The Internet of Things facilitates the collection of large amounts of data: sensors, smartphones, and even home appliances, generate a data deluge about individuals, their context and the events in their daily life. Providers can analyse these data in order to extract patterns and increase knowledge about their services, either on their own or by transferring datasets to third parties. To mitigate the Big Brother effect, i.e. to preserve the individuals’ right to privacy, techniques in the scope of Statistical Disclosure Control (SDC) must be applied. Microaggregation, is one of the best-known methods in the SDC arena. However, its results are far from optimal. In this paper, we introduce Random Cluster Shuffling, a new post-processing method that aims at improving the results of microaggregation techniques. We describe the proposal and present some results that support the potential of our approach.
Authors: M. Li, L. Zhu, Z. Zhang, C. Lal, M. Conti and M. Alazab
Date of Publication: April 2022