Skip to main content

Expert Meeting on Statistical Data Confidentiality

01 - 03 December 2021
Poznań Poland

The 2021 joint UNECE/Eurostat Expert Meeting on Statistical Data Confidentiality will be hosted by Statistics Poland in Poznań on 1-3 December 2021.

While this meeting is convened as an in-person event in Poland, for those participants who are unable to travel, we will try to provide a remote connection, upon request, to allow remote participation.

The focus of the meeting will be on cutting edge ideas, new trends, experiments, and approaches in the areas of statistical data confidentiality. In addition to the traditional presentations, the agenda of the meeting will include target-driven small group discussions and interactive activities. The programme of the meeting will aim to cover the following topics:

  • Access to microdata;
  • Microdata protection;
  • Tabular data;
  • Risk assessment: Privacy, confidentiality, and disclosure;
  • Other emerging issues;
  • Software tools for statistical data confidentiality; and
  • Communication of statistical disclosure control methods.

Documents

  Document Title Documents Presentations
  Provisional Programme

PDF

 
  Information Notice 1

PDF

 
  Information Notice 2 PDF  
  Report PDF  

Topic: Access to microdata

Session Organizers: Aleksandra Bujnowska (Eurostat) and Eric Schulte Nordholt (Statistics Netherlands)
  Access to different kinds of Statistics Netherlands’ microdata
Eric Schulte Nordholt (Statistics Netherlands)
Paper PDF
  Creating ready-made research datasets from national administrative registers
Päivi Kankaanranta (Statistics Finland)
Paper PDF
  Fingerprinting relational data
Tanja Šarčević (SBA Research)
Paper PDF
  Shedding light on the legal approach to aggregate data under the GDPR & the FFDR
Emanuela Podda (Università di Bologna)
Paper PDF
  Transnational access to confidential microdata: Progress and impact for research
Maria Alkhoury (Centre d'accès sécurisé aux données / Secure Data Hub)
Paper PDF
  Microdata access where we are and where we need to go
Elizabeth Green (University of the West of England)
Paper PDF
  Microdata access services coping with COVID-19 lockdown
Natalia Volkow (National Institute of Statistics and Geography)
Paper PDF

Topic: Risk assessment: Privacy, confidentiality, and disclosure

Session Organizers: Josep Domingo-Ferrer (Universitat Rovira i Virgili), Krish Muralidhar (University of Oklahoma)
  Statistical disclosure control for machine learning models
Felix Ritchie (University of the West of England)
Paper PDF
  The trade-off between the risk of disclosure and data utility in SDC – a case of data from a survey of accidents at work
Andrzej Młodak (Statistical Office in Poznań)
Paper PDF
  Using machine learning to assist output checking
Josep Domingo-Ferrer (Universitat Rovira i Virgili)
Paper PDF
  Disclosure metrics born from statistical evaluations of data utility
Devyani Biswal (University of Ottawa)
Paper PDF
  Risk assessment procedures for the 2020 U.S. census
David Van Riper (University of Minnesota)
Paper PDF
  Proposal for a risk assessment scale for privacy risks in the disclosure of statistical information
Jesús González López (National Institute of Statistics and Geography)
Paper PDF
  Database reconstruction is very difficult in practice
Krishnamurty Muralidhar (University of Oklahoma)
Paper PDF

Topic: Software tools for statistical data confidentiality

Session Organizers: Peter-Paul de Wolf (Statistics Netherlands) and Andrzej Młodak (Poznań Statistical Office)
  Suppression of directly-disclosive cells in frequency tables
Daniel Lupp (Statistics Norway)
Paper PDF
  Introducing a graphical user interface for creating the metadata governing the secondary cell suppression process
Michel Reiffert (Destatis)
Paper PDF
  Assessing, visualizing and improving the utility of synthetic data
Gillian Raab (Scottish Centre for Administrative Data Research)
Paper PDF
  Private linear regression: Can we scale up with Big Data?
Giuseppe Bruno (Bank of Italy)
Paper PDF
  Automatic checking of research outputs
Marco Stocchi (Eurostat)
Paper PDF

Topic: Microdata protection

Session Organizers: Josep Domingo-Ferrer (Universitat Rovira i Virgili), Krish Muralidhar (University of Oklahoma)
  Accounting for longitudinal data structures when disseminating synthetic data to the public
Joerg Drechsler (Institute for Employment Research)
Paper PDF
  AI-based privacy preserving census(like) data publication
Johannes Gussenbauer (Statistics Austria)
Paper PDF
  Generating tabular data using generative adversarial networks with differential privacy
Giacomo Astolfi (European Central Bank)
Paper PDF
  Generative adversarial networks for synthetic data generation: A comparative study
Claire Little (University of Manchester)
Paper PDF
  Data access modernization in National Statistical Offices through synthetic data, the HLG-MOS guide
Kenza Sallier (Statistics Canada)
  PDF
  Extreme value protection adjustment for different subpopulations in complex data sets
Anna Oganian (National Center for Health Statistics)
Paper PDF

Topic: Tabular data

Session Organizers: Steven Thomas (Statistics Canada) and Sarah Giessing (Destatis)
  Differential privacy and noisy confidentiality concepts for European population statistics
Fabian Bach (Eurostat)
Paper PDF
  Fair risk-utility comparison of tabular perturbation methods by post-processing to expected frequencies
Øyvind Langsrud (Statistics Norway)
Paper PDF
  Suppression or perturbation?
Wim Kloek (Eurostat)
Abstract PDF
  Considerations to deal with the frozen cell problem in Tau-Argus Modular
Sarah Giessing (Destatis)
Paper PDF
  Increasing utility of economic statistical information
Steven Thomas (Statistics Canada)
Paper PDF

Topic: Other Emerging issues

Session Organizers: Peter-Paul de Wolf (Statistics Netherlands) and Janika Tarkoma (Statistics Finland)
  Some of entity resolution and the impacts on personal privacy
Rebecca Steorts (Duke University)
Abstract PDF
  A proof-of-concept solution for secure processing of mobile network operator data for official statistics
Fabio Ricciato, (Eurostat)
Paper PDF
  Modelling data environments within PROV to assist anonymisation decision-making
Mark Elliot (University of Manchester)
Paper PDF
  Structural uniqueness in network data
Marieke de Vries (Statistics Netherlands)
Paper PDF

Updates on the HLG project on input privacy preservation 

  Dennis Ramondt (Statistics Netherlands)      PDF