Watermarking Generative Tabular Data
This work addresses security concerns for synthetic and real-world datasets, though it is incremental as it adapts existing watermarking concepts to tabular data.
The paper tackles the problem of watermarking generative tabular data by introducing a mechanism based on data binning and statistical hypothesis testing, achieving effective detection with theoretical guarantees while preserving data fidelity and robustness against noise attacks.
In this paper, we introduce a simple yet effective tabular data watermarking mechanism with statistical guarantees. We show theoretically that the proposed watermark can be effectively detected, while faithfully preserving the data fidelity, and also demonstrates appealing robustness against additive noise attack. The general idea is to achieve the watermarking through a strategic embedding based on simple data binning. Specifically, it divides the feature's value range into finely segmented intervals and embeds watermarks into selected ``green list" intervals. To detect the watermarks, we develop a principled statistical hypothesis-testing framework with minimal assumptions: it remains valid as long as the underlying data distribution has a continuous density function. The watermarking efficacy is demonstrated through rigorous theoretical analysis and empirical validation, highlighting its utility in enhancing the security of synthetic and real-world datasets.