Page 756

Ming Ding | Information Security and

Privacy Group | Data61, CSIROWhat can you do as an organization? Data synthesis

•Privacy preserved dataset: Creating a parallel universe, which hides everyone’s real life!

•A perturbed Gaussian Copula approach to generate data

–From the original dataset, we extract the PDF/CDF for each attribute 𝒊

–From the original dataset, we analyse the correlation matrix for all pairs of attributes (denoted by 𝑹)

–Then, we perturb the above statistics satisfying differential privacy and generate a privacy preserved dataset

https://www.scmp.com/news/china/diplomacy/article/217709

1/beijing -blames -canada -huawei -arrest -and-threatens -graveNo Photo

in a CourtroomSketch

Hassan Jameel Asghar, Ming Ding, Thierry Rakotoarivelo,

Sirine Mrabet, Dali Kaafar, “Differentially private release of

datasets using Gaussian copula,” Journal of Privacy and

Confidentiality, vol. 10, no. 2, pp. 1 -38, Jun. 2020.