An Empirical Study on the Membership Inference Attack against Tabular Data Synthesis Models
This addresses privacy risks for users of tabular data synthesis models, but it is incremental as it extends known attacks from image to tabular data.
The paper investigates membership inference attacks on tabular data synthesis models, finding that these attacks seriously jeopardize four state-of-the-art models, and shows that differentially-private training algorithms like DP-SGD and DP-GAN can largely alleviate the threat by sacrificing generation quality.
Tabular data typically contains private and important information; thus, precautions must be taken before they are shared with others. Although several methods (e.g., differential privacy and k-anonymity) have been proposed to prevent information leakage, in recent years, tabular data synthesis models have become popular because they can well trade-off between data utility and privacy. However, recent research has shown that generative models for image data are susceptible to the membership inference attack, which can determine whether a given record was used to train a victim synthesis model. In this paper, we investigate the membership inference attack in the context of tabular data synthesis. We conduct experiments on 4 state-of-the-art tabular data synthesis models under two attack scenarios (i.e., one black-box and one white-box attack), and find that the membership inference attack can seriously jeopardize these models. We next conduct experiments to evaluate how well two popular differentially-private deep learning training algorithms, DP-SGD and DP-GAN, can protect the models against the attack. Our key finding is that both algorithms can largely alleviate this threat by sacrificing the generation quality.