A social knowledge network for research

Cards

Collections

Explore

Search

Log in

Nick Vincent's cards — Semble

Nick Vincent

Nick Vincent

@nickmvincent.bsky.social

Studying people and computers (www.nickmvincent.com) Blogging about data and steering AI (dataleverage.substack.com)

6 Followers 0 Following 0 Collections Following

--- citation_key: staab2024verticaldataminimization entry_type: inproceedings title: 'From Principle to Practice: Vertical Data Minimization for Machine Learning' authors: - Robin Staab - Nikola Jovanović - Mislav Balunović - Martin Vechev year: '2024' venue: IEEE Symposium on Security and Privacy url: https://arxiv.org/abs/2311.10500 doi: 10.1109/SP54263.2024.00089 tags: - data minimization - machine learning - feature generalization - privacy --- Proposes vertical data minimization for ML through data generalization, reducing the need to collect full-resolution client data during training and deployment.

--- citation_key: pallas2022perquerydataminimization entry_type: inproceedings title: Configurable Per-Query Data Minimization for Privacy-Compliant Web APIs authors: - Frank Pallas - David Hartmann - Paul Heinrich - Josefine Kipke - Elias Grünewald year: '2022' venue: International Conference on Web Engineering url: https://arxiv.org/abs/2203.09903 doi: 10.1007/978-3-031-09917-5_22 tags: - data minimization - web APIs - GraphQL - privacy engineering --- Introduces Janus, a role-configurable GraphQL API layer for minimizing released personal data on a per-query basis.

--- citation_key: biega2020personalizationminimization entry_type: inproceedings title: Operationalizing the Legal Principle of Data Minimization for Personalization authors: - Asia J. Biega - Peter Potash - Hal Daumé III - Fernando Diaz - Michèle Finck year: '2020' venue: SIGIR url: https://arxiv.org/abs/2005.13718 doi: 10.1145/3397271.3401034 tags: - data minimization - personalization - recommender systems - GDPR --- Explores operational definitions of data minimization for personalization and recommender systems, including disparate impacts across users.

--- citation_key: pinisetty2018monitoringdataminimisation entry_type: misc title: Monitoring Data Minimisation authors: - Srinivas Pinisetty - Thibaud Antignac - David Sands - Gerardo Schneider year: '2018' venue: arXiv url: https://arxiv.org/abs/1801.02484 doi: 10.48550/arXiv.1801.02484 tags: - data minimization - runtime monitoring - formal methods --- Studies when data-minimization violations can be monitored at runtime for deterministic programs, including monolithic and distributed input settings.

--- citation_key: antignac2017languagebased entry_type: inproceedings title: 'Data Minimisation: A Language-Based Approach' authors: - Thibaud Antignac - David Sands - Gerardo Schneider year: '2017' venue: IFIP International Conference on ICT Systems Security and Privacy Protection url: https://arxiv.org/abs/1611.05642 doi: 10.1007/978-3-319-58469-0_30 tags: - data minimization - programming languages - formal methods --- Defines a data minimizer as a program preprocessor that reduces input information while preserving the purpose embodied by the program.

--- citation_key: veeningen2014communicationprotocols entry_type: article title: >- Data Minimisation in Communication Protocols: A Formal Analysis Framework and Application to Identity Management authors: - Meilof Veeningen - Benne de Weger - Nicola Zannone year: '2014' venue: International Journal of Information Security url: https://arxiv.org/abs/1206.7111 doi: 10.1007/s10207-014-0235-z tags: - data minimization - communication protocols - identity management - formal methods --- A formal framework for comparing communication protocols by the amount of personal information they expose, with identity-management case studies.

--- citation_key: staab2026sokdataminimization entry_type: inproceedings title: 'SoK: Data Minimization in Machine Learning' authors: - Robin Staab - Nikola Jovanović - Kimberly Mai - Prakhar Ganesh - Martin Vechev - Ferdinando Fioretto - Matthew Jagielski year: '2026' venue: IEEE Conference on Secure and Trustworthy Machine Learning url: https://arxiv.org/abs/2508.10836 doi: 10.48550/arXiv.2508.10836 tags: - data minimization - machine learning - systematization - privacy --- A systematization of data minimization in ML, organizing minimization points, adversarial models, and DM-adjacent methods under a common framework.

--- citation_key: shaowang2026iotdataminimization entry_type: article title: >- Algorithmic Data Minimization for Machine Learning over Internet-of-Things Data Streams authors: - Ted Shaowang - Shinan Liu - Jonatas Marques - Nick Feamster - Sanjay Krishnan year: '2026' venue: PVLDB url: https://arxiv.org/abs/2503.05675 doi: 10.14778/3773731.3773740 tags: - data minimization - machine learning - IoT - data streams --- Gives a technical interpretation of data minimization for sensor streams and evaluates methods that reduce identifiability while maintaining predictive accuracy.

--- citation_key: ganesh2024dataminimizationprinciple entry_type: inproceedings title: The Data Minimization Principle in Machine Learning authors: - Prakhar Ganesh - Cuong Tran - Reza Shokri - Ferdinando Fioretto year: '2024' venue: Regulatable ML Workshop at NeurIPS url: https://arxiv.org/abs/2405.19471 doi: 10.48550/arXiv.2405.19471 tags: - data minimization - machine learning - privacy - optimization --- Introduces an optimization framework for operationalizing legal data-minimization definitions and probes the gap between minimization and actual privacy protection.

--- citation_key: yao2024privacyqte entry_type: inproceedings title: >- Privacy-Preserving Quantile Treatment Effect Estimation for Randomized Controlled Trials authors: - Leon Yao - Paul Yiming Li - Jiannan Lu year: '2024' venue: Conference on Digital Experimentation url: https://arxiv.org/abs/2401.14549 doi: 10.48550/arXiv.2401.14549 tags: - data minimization - experimentation - differential privacy - causal inference --- Extends data-minimizing experimentation to quantile treatment effects via histogram aggregation, with an optional differential privacy layer.

--- citation_key: tran2023inferenceminimization entry_type: inproceedings title: Data Minimization at Inference Time authors: - Cuong Tran - Ferdinando Fioretto year: '2023' venue: NeurIPS url: >- https://proceedings.neurips.cc/paper_files/paper/2023/hash/e48880ea81caa7836e6a0694049093ae-Abstract-Conference.html tags: - data minimization - machine learning - inference - feature disclosure --- Studies personalized inference-time feature disclosure, asking whether each person can reveal only a small subset of attributes without losing decision accuracy.

--- citation_key: witt2023topicsapi entry_type: article title: >- The Interplay Between Machine Learning and Data Minimization Under the GDPR: The Case of Google's Topics API authors: - Cornelius Witt - Jan De Bruyne year: '2023' venue: International Data Privacy Law url: https://academic.oup.com/idpl/article/13/4/284/7425689 doi: 10.1093/idpl/ipad020 tags: - data minimization - GDPR - machine learning - advertising --- A legal analysis of how GDPR data minimization interacts with machine learning systems, using Google's Topics API as a concrete case.

--- citation_key: shanmugam2022limitdatacollection entry_type: inproceedings title: >- Learning to Limit Data Collection via Scaling Laws: A Computational Interpretation for the Legal Principle of Data Minimization authors: - Divya Shanmugam - Fernando Diaz - Samira Shabanian - Michele Finck - Asia J. Biega year: '2022' venue: FAccT url: https://arxiv.org/abs/2107.08096 doi: 10.1145/3531146.3533148 tags: - data minimization - scaling laws - data collection - GDPR --- Introduces FIDO, a framework that uses learned performance curves and diminishing returns to decide when additional personal-data collection is no longer justified.

--- citation_key: chou2021rctwithoutretention entry_type: inproceedings title: Randomized Controlled Trials without Data Retention authors: - Winston Chou year: '2021' venue: Conference on Digital Experimentation url: https://arxiv.org/abs/2102.03316 doi: 10.48550/arXiv.2102.03316 tags: - data minimization - experimentation - retention - causal inference --- Shows how treatment-effect inference can work under stringent retention limits by recursively updating estimates and deleting or anonymizing individual records soon after collection.

--- citation_key: goldsteen2022dataminimizationgdpr entry_type: article title: Data Minimization for GDPR Compliance in Machine Learning Models authors: - Abigail Goldsteen - Gilad Ezov - Ron Shmelkin - Micha Moffie - Ariel Farkash year: '2022' venue: AI and Ethics url: https://arxiv.org/abs/2008.04113 doi: 10.1007/s43681-021-00095-8 tags: - data minimization - machine learning - GDPR - feature generalization --- One of the earliest explicit ML treatments of GDPR data minimization, reducing or generalizing runtime input features while preserving model utility.

--- citation_key: senarath2019dataminimizationmodel entry_type: article title: A Data Minimization Model for Embedding Privacy into Software Systems authors: - Awanthika Senarath - Nalin Asanka Gamagedara Arachchilage year: '2019' venue: Computers & Security url: https://doi.org/10.1016/j.cose.2019.101605 doi: 10.1016/j.cose.2019.101605 tags: - data minimization - privacy engineering - software engineering --- A privacy-engineering methodology for helping developers reason about collection, storage, and sharing decisions across a system design.

From Principle to Practice: Vertical Data Minimization for Machine Learning

Aiming to train and deploy predictive models, organizations collect large amounts of detailed client data, risking the exposure of private information in the event of a breach. To mitigate this,...

Configurable Per-Query Data Minimization for Privacy-Compliant Web APIs

The purpose of regulatory data minimization obligations is to limit personal data to the absolute minimum necessary for a given context. Beyond the initial data collection, storage, and...

Operationalizing the Legal Principle of Data Minimization for...

Article 5(1)(c) of the European Union's General Data Protection Regulation (GDPR) requires that "personal data shall be [...] adequate, relevant, and limited to what is necessary in relation to...

Monitoring Data Minimisation

Data minimisation is a privacy enhancing principle, stating that personal data collected should be no more than necessary for the specific purpose consented by the user. Checking that a program...

Data Minimisation: a Language-Based Approach (Long Version)

Data minimisation is a privacy-enhancing principle considered as one of the pillars of personal data regulations. This principle dictates that personal data collected should be no more than...

Data Minimisation in Communication Protocols: A Formal Analysis...

With the growing amount of personal information exchanged over the Internet, privacy is becoming more and more a concern for users. One of the key principles in protecting privacy is data...

SoK: Data Minimization in Machine Learning

Data minimization (DM) describes the principle of collecting only the data strictly necessary for a given task. It is a foundational principle across major data protection regulations like GDPR...

Algorithmic Data Minimization for Machine Learning over...

Machine learning can analyze vast amounts of data generated by IoT devices to identify patterns, make predictions, and enable real-time decision-making. By processing sensor data, machine learning...

The Data Minimization Principle in Machine Learning

The principle of data minimization aims to reduce the amount of data collected, processed or retained to minimize the potential for misuse, unauthorized access, or data breaches. Rooted in...

Privacy-preserving Quantile Treatment Effect Estimation for...

In accordance with the principle of "data minimization", many internet companies are opting to record less data. However, this is often at odds with A/B testing efficacy. For experiments with...

proceedings.neurips.cc

Data Minimization at Inference Time

In high-stakes domains such as legal, banking, hiring, and healthcare, learning models frequently rely on sensitive user information for inference, necessitating the complete set of features. This not only poses significant privacy risks for individuals but also demands substantial human effort from organizations to verify information accuracy. This study asks whether it is necessary to use all input features for accurate predictions at inference time. The paper demonstrates that, in a personalized setting, individuals may only need to disclose a small subset of features without compromising decision-making accuracy. The paper also provides an efficient sequential algorithm to determine the appropriate attributes for each individual to provide. Evaluations across various learning tasks show that individuals can potentially report as little as 10\% of their information while maintaining the same accuracy level as a model that employs the full set of user information.

academic.oup.com

The interplay between machine learning and data minimization under the GDPR: the case of Google’s topics API

The rapid speed of digitalization and the continuous technological disruption have lead to different paradoxes and contradictions. With the widespread adop

https://academic.oup.com/idpl/article/13/4/284/7425689 social preview image

Learning to Limit Data Collection via Scaling Laws: A...

Modern machine learning systems are increasingly characterized by extensive personal data collection, despite the diminishing returns and increasing societal costs of such practices. Yet, data...

Randomized Controlled Trials without Data Retention

Amidst rising appreciation for privacy and data usage rights, researchers have increasingly acknowledged the principle of data minimization, which holds that the accessibility, collection, and...

Data Minimization for GDPR Compliance in Machine Learning Models

The EU General Data Protection Regulation (GDPR) mandates the principle of data minimization, which requires that only data necessary to fulfill a certain purpose be collected. However, it can...