PROGRAMME
Notes
- Presentations marked with *: Short presentation. Paper in usb proceedings
- Presentations should last 20 minutes + 5 minutes for questions. Short presentations: 10 + 5
- Download a PDF Version
Please, upload your slides to the room laptop before your session.
Wednesday, 25 September
| 08h30 | Registration |
| 09h15 | Opening Session. Melek Önen and Josep Domingo-Ferrer |
| Session: Disclosure risk assessment I. Session Chair: Marieke de Vries | |
| 09:20 | "The Statbarn: A New Model for Output Statistical Disclosure Control", Elizabeth Green, Felix Ritchie and Paul White |
| 09:45 | "Attribute Disclosure Risk in Smart Meter Data: A Position Paper", Guillermo Navarro-Arribas and Vicenç Torra |
| 10:10 | "An Examination of the Alleged Privacy Threats of Confidence-Ranked Reconstruction of Census Microdata", David Sanchez, Najeeb Jebreel, Krish Muralidhar, Josep Domingo-Ferrer and Alberto Blanco-Justicia |
| 10:35 | BREAK |
| Session: Disclosure risk assessment II. Session Chair: Matthias Templ | |
| 10:55 | "Synthetic Data: Comparing Utility and Risk in Microdata and Tables", Simon Xi Ning Kolb, Jui Andreas Tang and Sarah Giessing |
| 11:20 | "Synthetic Data Outliers: Navigating Identity Disclosure", Carolina Trindade, Luís Antunes, Tânia Carvalho and Nuno Moniz |
| 11:45 | "Privacy Risk from Synthetic Data: Practical Proposals", Gillian Raab |
| 12:10 | LUNCH BREAK |
| Session: Privacy models and concepts. Session Chair: Peter-Paul de Wolf | |
| 14:00 | "From Isolation to Identification", Giuseppe D'Acquisto, Aloni Cohen, Maurizio Naldi and Kobbi Nissim |
| 14:25 | "Differentially Private Quantile Regression", Tran Tran, Matthew Reimherr and Aleksandra Slavkovic |
| 14:50 | "Utility Analysis of Differentially Private Anonymized Data Based on Random Sampling", Takumi Sugiyama, Hiroto Osugi, Io Yamanaka and Kazuhiro Minami |
| 15:15 | "Privacy- & Utility-Preserving Data Releases over Fragmented Data Using Differential Privacy via Individual Ranking", Luis Del Vasto Terrientes, Sergio Martinez and David Sanchez |
| 15:30 | BREAK |
| Session: Statistical table protection. Session Chair: Sarah Giessing | |
| 15:50 | "Secondary Cell Suppression by Gaussian Elimination: An Algorithm Suitable for Handling Issues with Zeros and Singletons", Øyvind Langsrud |
| 16:15 | "Obtaining (ɛ,δ)-Differential Privacy Guarantees When Using the Poisson Distribution to Synthesize Tabular Data", James Jackson, Robin Mitra, Brian Francis and Iain Dove |
| Session: Microdata protection. Session Chair: Anna Oganian | |
| 16:40 | "Asymptotic Utility of Spectral Anonymization", Katariina Perkonoja and Joni Virta |
| 17:05 | "Robin Hood: A De-identification Method to Preserve Minority Representation for Disparities Research", James Thomas Brown, Ellen Clayton, Michael Matheny, Murat Kantarcioglu, Yevgeniy Vorobeychik and Bradley Malin |
| 17:30 | "An Optimization Approach to Privacy Preserving Dynamic Data Publishing", Jordi Castro, Claudio Gentile and Adrian Tobar-Nicolau * |
Thursday, 26 September
| Session: Synthetic data generation methods I. Session Chair: Paul Francis | |
| 09:00 | "The Production of Bespoke Synthetic Teaching Datasets without Access to the Original Data", Mark Elliot, Claire Little and Richard Allmendinger |
| 09:25 | "Evaluating the Pseudo Likelihood Approach for Synthesizing Surveys under Informative Sampling", Anna Oganian, Joerg Drechsler and Mehtab Iqbal |
| 09:50 | "Hidden Power of Quasi-multinomial Sampling: Utility Analysis and Bias Correction", Hajime Ono and Nobuaki Hoshino * |
| 10:05 | "Evaluation of Synthetic Data Quality Using the Quantile at Risk", Michel Béra, Vasiliki Daskalaki, Spiros Kolovos, Antonis Spinakis, Konstantinos Spinakis and Photis Stavropoulos * |
| 10:20 | "A Cautionary Reflection on (Pseudo-)Synthetic Data from Deep Learning on Personal Data", Fabio Ricciato * |
| 10:35 | BREAK |
| Session: Synthetic data generation methods II. Session Chair: Bradley Malin | |
| 10:55 | "Generating Synthetic Data Is Complicated: Know Your Data and Know Your Generator", Jonathan Latner, Marcel Neunhoeffer and Jörg Drechsler |
| 11:20 | "Developing Synthetic Microdata through Machine Learning for the Annual Business Survey", Jorge Cisneros, Audrey Kindlon, Timothy Wojan, Matthew Williams, Jennifer Ozawa, Christine Task, Damon Streat and Heather Madray * |
| Session: Synthetic data generation software. Session Chair: Gillian Raab | |
| 11:35 | "A Comparison of SynDiffix Multi-table versus Single-table Synthetic Data", Paul Francis |
| 12:00 | "An Evaluation of Synthetic Data Generators Implemented in the Python Library Synthcity", Emma Fössing and Jörg Drechsler |
| 12:25 | "Evaluation of Synthetic Data Generators on Complex Tabular Data", Oscar Thees, Jiří Novák and Matthias Templ |
| 12:50 | LUNCH BREAK |
| Session: Case studies. Session Chair: Jörg Drechsler | |
| 14:45 | "A Case Study Exploring Data Synthesis Strategies on Tabular vs. Aggregated Data Sources for Official Statistics", Mohamed Aghaddar, Liu Nuo Su, Manel Slokom and Peter-Paul de Wolf |
| 15:10 | "Relational Or Single: A Comparative Analysis of Data Synthesis Approaches for Privacy and Utility on a Use Case from Statistical Office", Shruti Agrawal, Manel Slokom, Nynke C. Krol and Peter-Paul de Wolf |
| 15:35 | "Escalation of Commitment: A Case Study of the United States Census Bureau Efforts to Implement Differential Privacy for the 2020 Decennial Census", Krish Muralidhar and Steven Ruggles |
| 16:00 | "Applications of Statistical Disclosure Control Methods to Protect the Confidentiality in Agricultural Census Microdata", Andrzej Młodak and Tomasz Józefowski * |
| 16:45 | Guide visit to Antibes Old Town |
| 19:30 | Gala Dinner at restaurant Le Café de la Plage |
Friday, 27 September
| Session: Spatial and georeferenced data. Session Chair: Krish Muralidhar | |
| 09:30 | "Masking Georeferenced Health Data - An Analysis Taking the Example of Partially Synthetic Data on Sleep Disorder", Simon Cremer, Lydia Jehmlich and Rainer Lenz |
| 09:55 | "Privacy and Disclosure Risks in Spatial Dynamic Microsimulations", Hanna Brenzel, Martin Palm, Jan Weymeirsch and Ralf Münnich |
| Session: Machine Learning and privacy I. Session Chair: Jordi Castro | |
| 10:20 | "Combinations of AI Models and XAI Metrics Vulnerable to Record Reconstruction Risk", Ryotaro Toma and Hiroaki Kikuchi |
| 10:45 | "DISCOLEAF: Personalized DIScretization of COntinuous Attributes for LEArning with Federated Decision Trees", Saloni Kwatra and Vicenc Torra |
| 11:10 | BREAK |
| Session: Machine Learning and privacy II. Session Chair: Sébastien Gambs | |
| 11:30 | "Node Injection Link Stealing Attack", Oualid Zari, Javier Parra-Arnau, Ayşe Ünsal and Melek Önen |
| 11:55 | "Assessing the Potentials of LLMs and GANs as State-of-the-art Tabular Synthetic Data Generation Methods", Marko Miletic and Murat Sariyar |
| 12:20 | "Active Learning for Human Annotation of Privacy-Preserved Synthetic Data", Osamu Saisho, Takayuki Miura, Kazuki Iwahana, Masanobu Kii and Rina Okada * |
| 12:35 | "Improving Utility in a DP-Fied ML Algorithm with a Metaheuristics-Based Privacy Budget Allocation Strategy", Marianne Abi Kanaan, Jean-François Couchot, Talar Atechian and Rony Darazi * |
| 12:50 | Closing remarks. Melek Önen and Josep Domingo-Ferrer |
Important Dates
- Submission deadline
MAY 19, 2024
MAY 26, 2024 - Acceptance notification June 21, 2024
- Proceedings version due June 30, 2024
- USB-only submission deadline June 30, 2024
- USB-only acceptance notification July 11, 2024
- USB-only proceedings version due July 17, 2024
