Projects

What may be released? Three current projects, one contribution and a selection of completed work since 2006.

Current

leading and contributing
  1. 2023 to 2027

    Privacy-preserving technologies for event and longitudinal data

    SNSF and Innosuisse, Bridge Discovery, grant 211751 · Principal investigator · Bern University of Applied Sciences BFH (Murat Sariyar) · CHF 710,208

    Mobility traces, life-course events and health records identify people through the sequence of what happened, so removing names is not enough. The project develops methods and open software to anonymise or synthesise such longitudinal data and to measure the risk that remains. Three strands run in parallel, the masking of location and movement data, the protection of event histories with time-varying attributes, and new risk measures beyond k-anonymity. Methods are tested on real data such as the Swiss TREE panel and published in the R packages sdcMicro, simPop and riskutility. SwissAnon grew out of this project.

    Grant on data.snf.ch · Project page at SwissAnon

    From event data to shareable data: Mobility traces, life-course events and health records identify people through the sequence of events, then Masking of location data, protection of event histories, risk measures beyond k-anonymity, then Anonymised or synthetic longitudinal data with measured residual risk, then Open software (sdcMicro, simPop, riskutility), tested on the TREE panel dataMobility traces, life-course events and health recordsidentify people through the sequence of eventsmethodMasking of location data, protection of event histories,risk measures beyond k-anonymityresultAnonymised or synthetic longitudinal data with measuredresidual riskuseOpen software (sdcMicro, simPop, riskutility), tested on theTREE panel
    From event data to shareable data
  2. 2026 to 2028

    Privacy-preserving data sharing for the SBB synthetic population

    SBB Research Fund · Principal investigator · SBB passenger planning, Federal Statistical Office as liaison

    The SBB synthetic population describes about nine million persons, four million households and 700,000 businesses with exact coordinates and is a central basis for transport planning. Because it rests on confidential register data of the Federal Statistical Office, it can hardly be shared, not even between departments. The project develops a household-aware anonymisation of the register-derived variables, protects the home coordinates and adds a formal disclosure-risk assessment, so that SBB and the Federal Statistical Office can decide which version may be released at which resolution. It builds on a 2025 bachelor thesis for the canton of Ticino.

    Project page at SwissAnon

    Sharing a synthetic population without exposing the registers: Synthetic population: 9 million persons, 4 million households, 700,000 businesses with coordinates, then Household-aware anonymisation of register variables, protected home coordinates, then Formal disclosure-risk assessment per version and resolution, then SBB and the Federal Statistical Office decide what may be released dataSynthetic population: 9 million persons, 4 millionhouseholds, 700,000 businesses with coordinatesmethodHousehold-aware anonymisation of register variables,protected home coordinatesresultFormal disclosure-risk assessment per version and resolutionuseSBB and the Federal Statistical Office decide what may bereleased
    Sharing a synthetic population without exposing the registers
  3. 2026

    Risk-based anonymisation and pseudonymisation for Nestlé Switzerland

    Service agreement with Société des Produits Nestlé SA · Project lead, with Roman Müller and Oscar Thees · Nestlé Switzerland

    Nestlé Switzerland analyses individual-level data from internal studies and wants to do so within data-protection law. The project assesses how easily people in such data could be re-identified, agrees acceptable risk thresholds with Nestlé and applies anonymisation, or optimised pseudonymisation where utility demands it, so that the data stay useful. The assessment covers k-anonymity as well as identity, attribute and inference disclosure.

    Project page at SwissAnon

    Risk-based anonymisation in practice: Individual-level data from internal studies, then Re-identification risk: k-anonymity, identity, attribute and inference disclosure, then Agreed risk thresholds, anonymisation or optimised pseudonymisation, then Analyses within data-protection law, data still useful dataIndividual-level data from internal studiesmethodRe-identification risk: k-anonymity, identity, attribute andinference disclosureresultAgreed risk thresholds, anonymisation or optimisedpseudonymisationuseAnalyses within data-protection law, data still useful
    Risk-based anonymisation in practice
  4. 2025 to 2029

    Epidemic source detection on contact networks, from theory to practice

    SNSF project funding, grant 228527, led by Martin Sterchi (FHNW) · Co-applicant · FHNW School of Business · CHF 391,938

    Where did an outbreak start? The project brings methods for locating the source of an epidemic on contact networks from theory to practice, with statistics for uncertain and incomplete network data.

    Grant on data.snf.ch

    Where did an outbreak start? Contact networks, uncertain and incomplete, then Source detection: statistics for locating the origin of an epidemic, then Methods that work on real network data, from theory to practice dataContact networks, uncertain and incompletemethodSource detection: statistics for locating the origin of anepidemicresultMethods that work on real network data, from theory topractice
    Where did an outbreak start?

Completed

selected, since 2006

A selection. The full list of projects and grants is in the CV.

  1. 2021 to 2022

    Anonymisation of census data and training in statistical disclosure control

    World Bank · Consultant · World Bank

    Anonymisation of census microdata and courses on statistical disclosure control for statistical offices.

  2. 2021

    Anonymisation of data for the Inter-American Development Bank

    Inter-American Development Bank · Consultant · Inter-American Development Bank

    Anonymisation of survey data for release.

  3. 2021

    Anonymisation of smart-meter data

    EnerSuisse · Project lead · EnerSuisse

    Methods for sharing electricity consumption time series without exposing households.

  4. 2020 to 2022

    Fellowship of the Digital Initiative of the Canton of Zurich on anonymisation

    Canton of Zurich · Fellow · ZHAW

    Two years of work on anonymisation methods and tools for public administration and research data.

  5. 2020 to 2021

    Better de-identification using statistical anonymisation for Malawi population health data

    ESTHER Switzerland · Project lead · Malawi Epidemiology and Intervention Research Unit

    Start-up grant and workshop on anonymising population health surveillance data.

  6. 2020

    Anonymisation of SBB movement data

    SBB · Project lead · SBB

    First study on protecting trajectory data of rail passengers.

  7. 2020

    Anonymisation of trajectory data

    Swisscom · Project lead · Swisscom

    Geo-masking of mobile-network movement data.

  8. 2019

    Anonymisation of health-insurance data

    Helsana Gesundheitswissenschaften · Project lead · Helsana

    Disclosure-risk assessment and anonymisation of claims data for research use.

  9. 2016

    Redesign of the graphical interface of sdcMicro

    World Bank · Consultant · World Bank

    The interface became sdcApp, the point-and-click front end of sdcMicro that statistical offices use to anonymise microdata.

    sdcMicro on CRAN

  10. 2015 to 2016

    Methods and tools for synthetic populations and synthetic data for microsimulation

    World Bank · Consultant · World Bank

    Two grants that produced the methods and the simPop package for building synthetic populations from survey and census data, used for microsimulation and for sharing data that cannot be released as they are.

    simPop paper

  11. 2015

    Skills gap survey in Nigeria

    UNIDO · Consultant · UNIDO

    Survey design, weighting and analysis of the skills gap survey for UNIDO in Nigeria.

  12. 2014 to 2015

    Improving the quality of household expenditure data and the reliability of poverty and inequality measures

    World Bank · Consultant · World Bank

    Detection and treatment of errors, outliers and missing values in household expenditure surveys so that poverty and inequality measures rest on cleaner data.

  13. 2009 to 2018

    Three projects with UNIDO: data editing, R in the statistics unit, energy statistics

    UNIDO · Consultant · UNIDO

    Editing and imputation of industrial statistics, the introduction of R in the statistics unit, and methods for energy statistics.

  14. 2008 to 2013

    Statistical disclosure control for the OECD and the International Household Survey Network

    OECD, IHSN · Consultant · OECD, IHSN

    Projects on disclosure control that shaped sdcMicro and its documentation for statistical offices.

    sdcMicro at the IHSN

  15. 2006 to 2016

    Tools and public-use files for statistical disclosure control in the European Statistical System

    Eurostat and the European Commission · Researcher and project lead · Statistics Austria and partners of the ESSnet on SDC

    A series of grants, from the Centre of Excellence in Statistical Disclosure Control (2006) and the ESSnet on SDC (2008 to 2011) to public-use files for EU-SILC and the Labour Force Survey (2014 to 2015) and tools for disclosure control (2016). The AMELI project of the 7th Framework Programme (2008 to 2011) worked on robust indicators of poverty and social exclusion.