| Abstract |
I present some ongoing work in collaboration with Luc Moreau and Kieron O'Hara at the University of Southampton. The police.uk website publishes a collection of open datasets of UK crime statistics. Before being published, data are smoothed for privacy reasons (in order not to cause distress to the people involved in such crimes). The goal of this work is twofold. First, we analyze the impact of different smoothing policies to the reliability of the resulting open datasets. Second, we compare the open datasets each other in order to derive hints about their reliability without having to refer to the closed raw data. I present some preliminary results and I outline the future of this research. |