Nicolas Grislain

Nicolas Grislain, Paul Roussel, Victoria de Sainte Agathe

This paper introduces Qrlew, an open source library that can parse SQL queries into Relations -- an intermediate representation -- that keeps track of rich data types, value ranges, and row ownership; so that they can easily be rewritten into differentially-private equivalent and turned back into SQL queries for execution in a variety of standard data stores. With Qrlew, a data practitioner can express their data queries in standard SQL; the data owner can run the rewritten query without any technical integration and with strong privacy guarantees on the output; and the query rewriting can be operated by a privacy-expert who must be trusted by the owner, but may belong to a separate organization.

Nicolas Grislain

Retrieval-Augmented Generation (RAG) has emerged as the dominant technique to provide \emph{Large Language Models} (LLM) with fresh and relevant context, mitigating the risk of hallucinations and improving the overall quality of responses in environments with large and fast moving knowledge bases. However, the integration of external documents into the generation process raises significant privacy concerns. Indeed, when added to a prompt, it is not possible to guarantee a response will not inadvertently expose confidential data, leading to potential breaches of privacy and ethical dilemmas. This paper explores a practical solution to this problem suitable to general knowledge extraction from personal data. It shows \emph{differentially private token generation} is a viable approach to private RAG.

Clément Sébastien Lalanne, Clément Gastaud, Nicolas Grislain et al.

We consider the differentially private estimation of multiple quantiles (MQ) of a distribution from a dataset, a key building block in modern data analysis. We apply the recent non-smoothed Inverse Sensitivity (IS) mechanism to this specific problem. We establish that the resulting method is closely related to the recently published ad hoc algorithm JointExp. In particular, they share the same computational complexity and a similar efficiency. We prove the statistical consistency of these two algorithms for continuous distributions. Furthermore, we demonstrate both theoretically and empirically that this method suffers from an important lack of performance in the case of peaked distributions, which can degrade up to a potentially catastrophic impact in the presence of atoms. Its smoothed version (i.e. by applying a max kernel to its output density) would solve this problem, but remains an open challenge to implement. As a proxy, we propose a simple and numerically efficient method called Heuristically Smoothed JointExp (HSJointExp), which is endowed with performance guarantees for a broad class of distributions and achieves results that are orders of magnitude better on problematic datasets.

Luca Canale, Nicolas Grislain, Grégoire Lothe et al.

In recent years, several models have improved the capacity to generate synthetic tabular datasets. However, such models focus on synthesizing simple columnar tables and are not useable on real-life data with complex structures. This paper puts forward a generic framework to synthesize more complex data structures with composite and nested types. It then proposes one practical implementation, built with causal transformers, for struct (mappings of types) and lists (repeated instances of a type). The results on standard benchmark datasets show that such implementation consistently outperforms current state-of-the-art models both in terms of machine learning utility and statistical similarity. Moreover, it shows very strong results on two complex hierarchical datasets with multiple nesting and sparse data, that were previously out of reach.

Nicolas Grislain, Joan Gonzalvez

The big-data revolution announced ten years ago does not seem to have fully happened at the expected scale. One of the main obstacle to this, has been the lack of data circulation. And one of the many reasons people and organizations did not share as much as expected is the privacy risk associated with data sharing operations. There has been many works on practical systems to compute statistical queries with Differential Privacy (DP). There have also been practical implementations of systems to train Neural Networks with DP, but relatively little efforts have been dedicated to designing scalable classical Machine Learning (ML) models providing DP guarantees. In this work we describe and implement a DP fork of a battle tested ML model: XGBoost. Our approach beats by a large margin previous attempts at the task, in terms of accuracy achieved for a given privacy budget. It is also the only DP implementation of boosted trees that scales to big data and can run in distributed environments such as: Kubernetes, Dask or Apache Spark.

Johan Leduc, Nicolas Grislain

Generative modeling has recently seen many exciting developments with the advent of deep generative architectures such as Variational Auto-Encoders (VAE) or Generative Adversarial Networks (GAN). The ability to draw synthetic i.i.d. observations with the same joint probability distribution as a given dataset has a wide range of applications including representation learning, compression or imputation. It appears that it also has many applications in privacy preserving data analysis, especially when used in conjunction with differential privacy techniques. This paper focuses on synthetic data generation models with privacy preserving applications in mind. It introduces a novel architecture, the Composable Generative Model (CGM) that is state-of-the-art in tabular data generation. Any conditional generative model can be used as a sub-component of the CGM, including CGMs themselves, allowing the generation of numerical, categorical data as well as images, text, or time series. The CGM has been evaluated on 13 datasets (6 standard datasets and 7 simulated) and compared to 14 recent generative models. It beats the state of the art in tabular data generation by a significant margin.

Grégoire Jauvion, Nicolas Grislain

In this paper, we consider the problem of optimizing the revenue a web publisher gets through real-time bidding (i.e. from ads sold in real-time auctions) and direct (i.e. from ads sold through contracts agreed in advance). We consider a setting where the publisher is able to bid in the real-time bidding auction for each impression. If it wins the auction, it chooses a direct campaign to deliver and displays the corresponding ad. This paper presents an algorithm to build an optimal strategy for the publisher to deliver its direct campaigns while maximizing its real-time bidding revenue. The optimal strategy gives a formula to determine the publisher bid as well as a way to choose the direct campaign being delivered if the publisher bidder wins the auction, depending on the impression characteristics. The optimal strategy can be estimated on past auctions data. The algorithm scales with the number of campaigns and the size of the dataset. This is a very important feature, as in practice a publisher may have thousands of active direct campaigns at the same time and would like to estimate an optimal strategy on billions of auctions. The algorithm is a key component of a system which is being developed, and which will be deployed on thousands of web publishers worldwide, helping them to serve efficiently billions of ads a day to hundreds of millions of visitors.