Data and Algorithmic Transparency and Accountability
Inria Associate Team 2018-2020
Privatics, Inria, France - LATECE, UQAM, Canada

Data and Algorithmic Transparency and Accountability



The recent advent of personalized services, which are tailored according to the interests of their users, has lead to the massive collection of personal data and the construction of detailed profiles about these users. However, in the current Internet paradigm, there is a strong asymmetry in terms of power and information between the entities that gather and process personal data (e.g., major Internet companies, telecom operators, cloud providers, ...) and the individuals to whom those data are related. As a consequence, users of information technologies have no choice but to trust the entities collecting and using their data. This lack of transparency gives rise to ethical issues such as loss of control over personal data, discrimination, and unfair processing. For instance, controversial practices have been observed such as price discrimination (i.e., customizing prices for users according to their browsing history or their location), gender discrimination as well as price steering (i.e., changing the order of search results to highlight specific products). Unfair treatments can result from biases in the data or wrong predictions made by the machine learning process used to perform the personalization. In addition, the structure of these machine learning algorithms is often complex and producing intelligible explanations about their results is quite challenging. All these issues become even more critical when this type of system is used to help decision making (or to make decisions) in sectors such as justice or health.


To address these issues, data and algorithmic transparency is a growing and emerging research area. In addition, increasing the transparency of algorithms and opening them to the scrutiny of the public or to independent authorities such as the CNIL is only the first step in order to make them more accountable. In particular, once the opacity of personalization algorithms has been lifted, one of the long-term objective is to able to measure and reduce the discrimination in these algorithms. We believe that the data and algorithmic transparency and accountability should address three aspects: (1) the traceability of the collection of personal data; (2) the verification of the compliance of algorithms with critical properties (and legal requirements) such as non-discrimination and fairness, and (3) the capacity to explain the results of algorithmic systems. The proposed Associate Team will build on the expertise of the partners to address these three complementary aspects.


The main scientific objective of this collaboration is to be able to advance the research on data and algorithmic transparency and accountability by studying it in concrete but diverse contexts. The first context that we propose to investigate thoroughly is how personal information related to the physical world, such as the presence and mobility data of users, are collected and exploited by companies to perform behavioral targeting.


The challenges are multiple. First, cyber-physical tracking systems are often passively collecting data or use deceptive techniques, and are thus hard to detect. Second, the collection and analysis of data is also a challenging task as we need to capture enough data to be able to infer and model the link between the activities of the users (in particular their mobility) and the personalized content. Finally, even if enough data are available, inferring the behavior of an algorithm is not trivial as we may only be able to interact with it as a black-box and thus we may need to approximate its behavior following a machine learning approach.


To overcome these challenges, we will split the research program of the associated team into three phases. During the first phase, we have identified existing tracking systems and studies and reconstructed their underlying mechanisms. Then, during a second phase, we will design a methodology to quantify how well a particular system meets requirements such as non-discrimination and fairness. And finally, during the third phase, we will develop methods for improving the intelligibility of systems, for example through the generation of global explanations of the logic of the system and local explanations about specific results. We will put emphasis on the interactions with the users to enhance their understanding of the system.


We will then build on these results to develop broader mechanisms that could be applied to other areas such as predictive justice and health. The investigation of this area will be done in collaboration with researchers in law from Québec and with experts of algorthmic decision making in the medical sector (such as the designers and operators of algorithms used in France to take decisions about organ transplantations).


Team

Ulrich Matchi AÏvodji

Postdoc, LATECE research group, UQAM, Québec

Marc Queudot

Masters' Student, LATECE research group, UQAM, Québec

Louis Béziaud

PhD Student, LATECE research group, UQAM, Québec

Théo Jourdan

PhD Student, Privatics research group, Inria, France

Clément Henin

PhD Student, Privatics research group, Inria, France

Marie-Jean Meurs

Assistant Professor, LATECE research group, UQAM, Québec

Guillaume Célosia

PhD Student, INSA-Lyon, Inria, Privatics research group, Lyon, France

Sébastien Gambs

Associate Professor, LATECE research group, UQAM, Québec

Mathieu Cunche

Assistant Professor, INSA-Lyon, Inria, Privatics research group, Lyon, France

Daniel Le Métayer

Senior Researcher, Inria, Privatics research group, Lyon, France

Rosin Claude Ngueveu

PhD Student, LATECE research group, UQAM, Québec

Antoine Boutet

Assistant Professor, INSA-Lyon, Inria, Privatics research group, Lyon, France

Dissemination

  • Media

    • May 4th, 2019:
      Tout compte fait: souriez, vous êtes géolocalisés !
      Antoine Boutet, Mathieu Cunche
      Vidéo

    • May 27th, 2019:
      La face cachée des objets connectés
      Mathieu Cunche
      Link

    • September 21st, 2019:
      Interview in Atlantico about algorithmic decision systems
      Daniel Le Métayer
      Link

    • July 25th, 2019:
      Données personnelles: un secret mal gardé
      Sébastien Gambs
      Link

    • July 13th, 2019:
      OK Google, arrête de m'enregistrer
      Sébastien Gambs
      Link

  • Events

    • October - December, 2020:
      Challenge on data anonymisation
      Challenge organised between three different groups of students at Insa Lyon and Bourges - Antoine Boutet, Mathieu Cunche, Sébastien Gambs

    • December 3 rd, 2020:
      DYSAN: Dynamically sanitizing motion sensor data against sensitive inferences through adversarial networks.
      Keynote given at ICT4V Workshop on Privacy and Anonymization - Antoine Boutet.
      More information about this event

    • June 8 th, 2020:
      Traitement automatique du langage naturel applique au domaine juridique.
      Keynote given at TALN 2020, Nancy - Marie-Jean Meurs, Hugo Cyr.
      More information about this event

    • October 10th, 2019:
      2nd Workshop on Data Transparency
      This event organised at Lyon (France) included 6 speakers and attracted more than 70 attendees - Antoine Boutet
      More information about this workshop

    • September - November, 2019:
      Challenge on data anonymization
      Challenge organised between three different groups of students at Insa Lyon and Bourges - Antoine Boutet, Mathieu Cunche, Sébastien Gambs

    • September 11th, 2019:
      Atelier Transparence et responsabilité algorithmique : approches et défis
      Montréal, Québec - Marie-Jean Meurs, Sébastien Gambs, Antoine Boutet, Daniel Le Métayer, Clément Henin, Rosin Claude Ngueveu, Ulrich Matchi AÏvodji
      More information about this workshop

    • September 27th, 2019:
      How to promote a responsable design and usage of decision making systems?
      Panel at the inaugural session of the Internet and Society Center - Daniel Le Métayer, Alexis Tsoukias. Paris.

    • June 25th, 2019:
      Algorithmic decision making: risks and opportunities for society
      Interview about our report for the European Parliament - Claude Castelluccia, Daniel Le Métayer.

    • June 3rd, 2019:
      Building trust in AI
      Workshop organized during the Global Science Week at Lyon - Claude Castelluccia, Daniel Le Métayer.

    • Junuary 30th, 2019:
      "Influence or manipulation? What protections in the digital world?"
      Panel organized at the 12th CPDP (international Conference on Computers Privacy and Data Protection) - Daniel Le Métayer.

    • May 22th, 2019:
      Cybersécurité
      Panel at the Pint of Science - Mathieu Cunche.

    • June 13th, 2019:
      Mécanismes anti-tracage dans les réseaux sans-fil
      Panel at the plenary session of the GDR Sécu - Mathieu Cunche.

    • May 21th, 2019:
      L’impact des objets connecté dans le quotidien des familles.
      Panel at UNAF - Mathieu Cunche.

    • March 29th, 2019:
      Utilisation des données de transport et anonymat : deux visions antagonistes ?
      Panel at InOut - Mathieu Cunche.

    • October 18th, 2019:
      Intelligence artificielle et impacts sociaux.
      Ateliers SociologIA, Montréal, Québec - Marie-Jean Meurs.

    • October 9th, 2019:
      Journées droit et IA : L’intelligence artificielle et la personne
      Event organized at Montreal, Quebec - Marie-Jean Meurs.

    • November 27th, 2019:
      Intelligence Artificielle et éthique
      Panel, Avignon, France - Marie-Jean Meurs.

    • April 22nd, 2019:
      Fairwashing: the risk of rationalization
      Seminar talk, Riken AIP, Tokyo, Japan - Sébastien Gambs.

    • November 29th, 2018:
      Respect de la vie privée et problématiques éthiques a l'ere des données massives
      Seminar talk, école IVADO/RALI en analyse du langage naturel - Sébastien Gambs.

    • November 15th, 2018:
      Respect de la vie privée et problématiques éthiques a l’ere des données massives
      Seminar du CRIM - Sébastien Gambs.

    • November 15th, 2018:
      Privacy and Ethical Issues in Big Data: Current Trends and Future Challenges
      Keynote talk, FPS 2018 - Sébastien Gambs.

    • January 24th, 2018:
      Panel on "Physical tracking: nowhere to hide?"
      Panel organised at CPDP (11th international Conference on Computers Privacy and Data Protection), Brussels, Belgium - Daniel Le Métayer, Mathieu Cunche
      Video

    • March 5-8, 2018:
      Shonan Meeting on Anonymization methods and inference attacks: theory and practice
      This event included multiple experts in the field, Shonan Village, Japan - Sébastien Gambs, Antoine Boutet
      More information about this event

    • April 23th, 2018:
      1st Workshop on Data Transparency
      This event organised at Lyon (France) included 9 speakers and more than 40 attendees - Antoine Boutet, Daniel Le Métayer, Sébastien Gambs
      More information about this workshop

    • May 18th, 2018:
      Journée thématique "Intelligibilité et Transparence du Machine Learning et des Intelligences Artificielles"
      Transparence, intelligibilité des algorithmes : quels besoins ? Quels moyens ?
      Conférence organisée par la SFdS (Société Française de Statistique), Paris, France - Daniel Le Métayer

    • November 15th, 2018:
      Panel on "Cybersécurité : se préparer au Tsunami"
      Sommet des Start-up Challenge, Lyon - Mathieu Cunche
      More information about this event

    • November 21th, 2018:
      Colloque HumanIA 2018, Montréal, Quebec
      This event included multiple speakers and tutorials - Marie-Jean Meurs, Sébastien Gambs, Mathieu Cunche, Antoine Boutet, Guillaume Célosia, Rosin Claude Ngueveu, Louis Béziaud
      More information about this event

    • October - December, 2018:
      Challenge on data anonymisation
      Challenge organised between three different groups of students at Insa Lyon and Bourges - Antoine Boutet, Mathieu Cunche, Sébastien Gambs

  • Publications

    • Antoine Boutet, Mathieu Cunche, Sébastien Gambs, Antoine Laurent, Benjamin Nguyen
      DARC: Data Anonymization and Re-identification Challenge.
      Paper accepted to RESSI 2020, December 2020.

    • Antoine Boutet, Carole Frindel, Sébastien Gambs, Theo Jourdan, Claude Rosin Ngueveu.
      DYSAN: Dynamically sanitizing motion sensor data against sensitive inferences through adversarial networks.
      ArXiv preprint arXiv:2003.10325 (2020).

    • Marco Fiore, Panagiota Katsikouli, Elli Zavou, Mathieu Cunche, Francoise Fessant, Dominique Le Hello, Ulrich Aivodji, Baptiste Olivier, Tony Quertier, Razvan Stanica.
      Privacy in trajectory micro-data publishing: a survey.
      Paper published in Transactions on Data Privacy, IIIA-CSIC, 2020, 13, pp.91 - 149.

    • Guillaume Celosia, Mathieu Cunche.
      Discontinued Privacy: Personal Data Leaks in Apple Bluetooth-Low-Energy Continuity Protocols.
      Proceedings on Privacy Enhancing Technologies, De Gruyter Open, 2020, 2020, pp.26 - 46.

    • Guillaume Celosia, Mathieu Cunche.
      Valkyrie: A Generic Framework for Verifying Privacy Provisions in Wireless Networks.
      WiSec 2020 - 13 th ACM Conference on Security and Privacy in Wireless and Mobile Networks, Jul 2020, Linz, Austria.

    • Clément Henin, Daniel Le Métayer.
      A Multi-layered Approach for Tailored Black-box Explanations.
      Proceedings of the Explainable Deep Learning-AI workshop (EDL-AI 2020), in conjunction with ICPR2020, the 25th International Conference on Pattern Recognition, Springer, LNCS.

    • Sonia Desmoulin-Canselier, Daniel Le Metayer.
      Décider avec les algorithmes - Quelle place pour l'Homme, quelle place pour le droit ?
      Dalloz - Les Sens Du Droit - February 2020.

    • Clément Henin, Daniel Le Métayer.
      A Generic Framework for Black-box Explanations.
      International Workshop on Fair and Interpretable Learning Algorithms (FILA 2020), in conjunction with the IEEE International Conference on Big Data (IEEE BigData 2020).

    • Clément Henin, Daniel Le Métayer.
      Towards a framework for challenging ML-based decisions.
      DeceptECAI 2020 - 1st International Workshop on Deceptive AI @ECAI2020, Aug 2020.

    • Tristan Allard, Louis Béziaud, Sébastien Gambs.
      Online publication of court records: circumventing the privacy-transparency trade-off.
      ArXiv preprint arXiv:2007.01688 (2020).

    • Tristan Allard, Louis Béziaud, Sébastien Gambs, Julien Lolive.
      Ouvrir la boite noire des algorithmes de personnalisation.
      Le profilage en ligne: entre libéralisme et régulation.

    • Théo Jourdan, Antoine Boutet, Amine Bahi, Carole Frindel.
      Privacy- Preserving IoT framework for activity recognition in personal healthcare monitoring.
      ACM Transactions on Computing for Healthcare, 2020, 2 (1), 1-22.

    • Antoine Boutet, Mathieu Cunche.
      Privacy Protection for Wi-Fi Location Positioning Systems.
      Journal of information security and applications, 2020.

    • Noelie Debs, Theo Jourdan, Ali Moukadem, Antoine Boutet, Carole Frindel.
      Motion sensor data anonymization by time-frequency filtering.
      2020 28th European Signal Processing Conference (EUSIPCO), 1707-1711.

    • Elham Mohammadi, Nada Naji, Louis Marceau, Marc Queudot, Eric Charton, Leila Kosseim, Marie-Jean Meurs.
      Cooking Up a Neural-based Model for Recipe Classification.
      Proceedings of The 12th Language Resources and Evaluation Conference, 5000-5009.

    • Marc Queudot, Eric Charton, Marie-Jean Meurs.
      Improving Access to Justice with Legal Chatbots.
      Stats 2020, 3 (3), 356-375.

    • Ulrich Aivodji, Alexandre Bolot, Sebastien Gambs.
      Model extraction from counterfactual explanations.
      ArXiv preprint arXiv:2009.01884 (2020).

    • Antoine Boutet, Sébastien Gambs
      Inspect what your location history reveals about you - Raising user awareness on privacy threats associated with disclosing his location data
      Demonstration published at CIKM 2019, November 2019, Beijing.

    • Antoine Boutet, Carole Frindel, Sébastien Gambs, Théo Jourdan, Rosin Claude Ngueveu
      Protecting motion sensor data against sensitive inferences through an adversarial network approach
      Paper accepted to APVP 2019, July 2019.

    • Clément Hénin, Daniel Le Métayer
      Towards a generic framework for black-box explanation methods
      IJCAI 2019 Workshop on Explainable Artificial Intelligence (XAI), August 2019. Inria Research Report (extended version).

    • Jérémie Decouchant, Antoine Boutet, Jiangshan Yu, Paulo Esteves-Verissimo
      P3LS: Plausible Deniability for Practical Privacy-Preserving Live Streaming
      SRDS 2019, November, Lyon, France.

    • Théo Jourdan, Antoine Boutet, Carole Frindel
      Vers la protection de la vie privée dans les objets connectés pour la reconnaissance d'activité en santé
      Technique et Science Informatiques (TSI) Journal, 2019.

    • Guillaume Céliosa, Mathieu Cunche
      Saving Private Addresses: An Analysis of Privacy Issues in the Bluetooth-Low-Energy Advertising Mechanism
      APVP 2019, July 2019.

    • Théo Jourdan, Antoine Boutet, Carole Frindel
      Activity recognition: keeping sensory data private by local model-based reinforcement learning
      APVP 2019, July 2019.

    • Ulrich Aivodji, Hiromi Arai, Olivier Fortineau, Sébastien Gambs, Satoshi Hara, Alain Tapp
      Fairwashing: the risk of rationalization
      ICML 219, June 2019, Long Beach, US.

    • Ulrich Matchi AÏvodji, Sébastien Gambs, Alexandre Martin
      IOTFLA : A Secured and Privacy-Preserving Smart Home Architecture Implementing Federated Learning
      IEEE Symposium on Security and Privacy Workshops 2019: 175-180

    • Mathieu Cunche, Daniel Le Métayer, Victor Morel
      A generic information and consent framework for the IoT
      18th IEEE International Conference On Trust, Security And Privacy In Computing and Communications (IEEE TrustCom 19), August 2019.

    • Raul Pardo, Daniel Le Métayer
      Analysis of privacy policies to enhance informed consent
      33rd Annual IFIP WG 11.3 Conference on Data and Applications Security and Privacy (DBSec 2019), July 2019. Inria Research Report (extended version).

    • Claude Castelluccia, Daniel Le Métayer
      Understanding algorithmic decision-making: Opportunities and challenges
      Study for the European Parliament (STOA), March 2019.

    • Yves-Alexandre de Montjoye, Sébastien Gambs, Vincent Blondel, Geoffrey Canright, Nicolas de Cordes, Sébastien Deletaille, Kenth Engø-Monsen, Manuel Garcia-Herranz, Jake Kendall, Cameron Kerry, Gautier Krings, Emmanuel Letouzé, Miguel Luengo-Oroz, Nuria Oliver, Luc Rocher, Alex Rutherford, Zbigniew Smoreda, Jessica Steele, Erik Wetter, Alex “Sandy” Pentland and Linus Bengtsson
      On the privacy-conscientious use of mobile phone data
      Scientific Data. Issue 5 (2018).

    • Ehab ElSalamouny and Sébastien Gambs
      Optimal noise functions for location privacy on continuous regions
      International Journal of Information Security: 613-630 (2018).

    • Sébastien Gambs
      Privacy and ethical challenges in Big Data (invited)
      Proceedings of the 11th International Symposium on Foundations and Practice of Security (FPS 2018), pp. 17-26.

    • Sébastien Gambs, Julien Lolive and Jean-Marc Robert
      Entwining Sanitization and Personalization on Databases
      Proceedings of the 2018 on Asia Conference on Computer and Communications Security (AsiaCCS 2018), pp 207-219.

    • Sonia Desmoulin-Canselier, Daniel Le Métayer
      Algorithmic decision systems in the health and justice sectors: certification and explanations for algorithms in European and French law
      European Journal of Law and Technology, Vol. 9, No 3, December 2018.

    • Antoine Briand, Sara Zacharie, Ludovic Jean-Louis, Marie-Jean Meurs
      Identification of Sensitive Content in Data Repositories to Support Personal Information Protection
      International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, Pages 898-910, 2018.

    • Antoine Boutet, Alexandre van Beurden, Sébastien Gambs
      Inspect what your location history reveals about you
      Demonstration given at OPERANDI 2018 (Open Day for Privacy, Transparency and Decentralization) collocated to PETS 2018, Barcelona, Spain, 2018.

    • Guillaume Celosia, Mathieu Cunche
      Detecting smartphone state changes through a Bluetooth based timing attack
      Paper published at WiSec 2018, Stockholm, Sweden, 2018.

    • Mohammad Alaggan, Mathieu Cunche, and Sébastien Gambs
      Privacy-preserving Wi-Fi Analytics
      Paper published at PETS 2018, Barcelona, Spain, 2018.

    • Théo Jourdan, Antoine Boutet, Carole Frindel
      Toward privacy in IoT mobile devices for activity recognition
      Paper published at Mobiquitous 2018, New York, United States.

    • Claude Castelluccia, Daniel Le Métayer
      Understanding algorithmic : opportunities and challenges
      Paper published by the European Parliament, Scientific Foresight Unit (STOA), 2018.

    • Sourya Joyee De, Daniel Le Métayer
      Privacy risk analysis to enable informed privacy settings
      Paper published at International Workshop on Privacy Engineering (IWPE), Euro S&P Workshops, 2018.

    • Sébastien Gambs, Julien Lolive, Jean-Marc Robert
      Entwining Sanitization and Personalization on Databases
      Paper published at Asia Conference on Computer and Communications Security (AsiaCCS), 2018.

    • Daniel Le Métayer, Pablo Rauzy
      Capacity: an abstract model of control over personal data
      Paper published at Conference on Data and Application Security and Privacy (CODASPY), 2018.

    • Marc Queudot, Marie-Jean Meurs
      Artificial Intelligence and Predictive Justice: Limitations and Perspectives
      Paper published at International Conference on Industrial, Engineering and Other Applications of Applied Intelligent Systems, 2018.

Contact Us

We will do our best to answer your request as soon as possible.