Mathias Humbert

Mathias Humbert, Ph.D. (EPFL, 2015)

I am an associate professor in the Department of Information Systems (Faculty of Business and Economics) at the University of Lausanne. Prior to this, I was a scientific project manager at the Cyber-Defence Campus, a senior data scientist at the Swiss Data Science Center (SDSC), ETHZ and EPFL, and a postdoctoral researcher at the Center for IT-Security, Privacy, and Accountability (CISPA), in Saarbrücken. I completed my Ph.D. thesis in 2015 in the School of Computer and Communication Sciences at EPFL (Switzerland), after B.Sc./M.Sc. studies at EPFL and UC Berkeley.

My research interests lie at the intersection of privacy, cybersecurity, and machine learning. I am currently investigating privacy risks in various contexts, such as machine learning as a service (MLaaS), online social networks, wearable devices, and spectrum monitoring. Furthermore, I am applying machine learning and data analytics to earlier detect cyber-threats. My research expertise includes genomic privacy, privacy in online social networks, location privacy, game theory, and probabilistic graphical models.

In my Ph.D. thesis (available here), I have notably analyzed the variety of interdependencies in individuals' data, and their impact on privacy. I have shown the impossibility of protecting privacy individually and independently of others (stemming from the lack of control over correlated data), and the need for new definitions of privacy, in its social context, and as a common good.

Main Publications (Google Scholar, DBLP)

Since 2016

Min Chen, Zhikun Zhang, Tianhao Wang, Michael Backes, Mathias Humbert, Yang Zhang. Graph Unlearning. In Proceedings of the 29th ACM Conference on Computer and Communication Security (CCS), 2022.
PDF BibTeX Abstract
```
@inproceedings{chen2022graph,
  title={Graph unlearning},
  author={Chen, Min and Zhang, Zhikun and Wang, Tianhao and Backes, Michael and Humbert, Mathias and Zhang, Yang},
  booktitle={Proceedings of the 2022 ACM SIGSAC Conference on Computer and Communications Security},
  pages={499--513},
  year={2022}
}
                
```
Machine unlearning is a process of removing the impact of some training data from the machine learning (ML) models upon receiving removal requests. While straightforward and legitimate, retraining the ML model from scratch incurs a high computational overhead. To address this issue, a number of approximate algorithms have been proposed in the domain of image and text data, among which SISA is the state-of-the-art solution. It randomly partitions the training set into multiple shards and trains a constituent model for each shard. However, directly applying SISA to the graph data can severely damage the graph structural information, and thereby the resulting ML model utility. In this paper, we propose GraphEraser, a novel machine unlearning framework tailored to graph data. Its contributions include two novel graph partition algorithms and a learning-based aggregation method. We conduct extensive experiments on five real-world graph datasets to illustrate the unlearning efficiency and model utility of GraphEraser. It achieves 2.06 (small dataset) to 35.94 (large dataset) unlearning time improvement. On the other hand, GraphEraser achieves up to 62.5% higher F1 score and our proposed learning-based aggregation method achieves up to 112% higher F1 score
Asuman Senol, Gunes Acar, Mathias Humbert, Frederik Zuiderveen Borgesius. Leaky Forms: A Study of Email and Password Exfiltration Before Form Submission. In Proceedings of the 31st USENIX Security Symposium, 2022..
PDF BibTeX Abstract
```
@inproceedings{senol22,
  title={Leaky Forms: A Study of Email and Password Exfiltration Before Form Submission},
  author={Senol, Asuman and Acar, Gunes and Humbert, Mathias and Borgesius, Frederik Zuiderveen},
  booktitle={31st $\{$USENIX$\}$ Security Symposium ($\{$USENIX$\}$ Security)},
  year={2022}
}

                
```
Web users enter their email addresses into online forms for a variety of reasons, including signing in or signing up for a service or subscribing to a newsletter. While enabling such functionality, email addresses typed into forms can also be collected by third-party scripts even when users change their minds and leave the site without submitting the form. Email addresses—or identifiers derived from them—are known to be used by data brokers and advertisers for cross-site, cross- platform, and persistent identification of potentially unsuspect- ing individuals. In order to find out whether access to online forms is misused by online trackers, we present a measure- ment of email and password collection that occurs before the form submission on the top 100,000 websites. We evaluate the effect of user location, browser configuration, and inter- action with consent dialogs by comparing results across two vantage points (EU/US), two browser configurations (desk- top/mobile), and three consent modes. Our crawler finds and fills email and password fields, monitors the network traffic for leaks, and intercepts script access to filled input fields. Our analyses show that users’ email addresses are exfiltrated to tracking, marketing and analytics domains before form submission and without giving consent on 1,844 websites in the EU crawl and 2,950 websites in the US crawl. While the majority of email addresses are sent to known tracking domains, we further identify 41 tracker domains that are not listed by any of the popular blocklists. Furthermore, we find incidental password collection on 52 websites by third-party session replay scripts.
Victor Cochard, Damian Pfammatter, Chi Thang Duong, Mathias Humbert. Investigating Graph Embedding Methods for Cross-Platform Binary Code Similarity Detection. In Proceedings of the 7th IEEE European Symposium on Security and Privacy (EuroS&P), 2022.
PDF BibTeX Abstract
```
@INPROCEEDINGS{9797388,
  author={Cochard, Victor and Pfammatter, Damian and Duong, Chi Thang and Humbert, Mathias},
  booktitle={2022 IEEE 7th European Symposium on Security and Privacy (EuroS&P)}, 
  title={Investigating Graph Embedding Methods for Cross-Platform Binary Code Similarity Detection}, 
  year={2022},
  volume={},
  number={},
  pages={60-73},
  doi={10.1109/EuroSP53844.2022.00012}}

                
```
IoT devices are increasingly present, both in the industry and in consumer markets, but their security remains weak, which leads to an unprecedented number of attacks against them. In order to reduce the attack surface, one approach is to analyze the binary code of these devices to early detect whether they contain potential security vulnerabilities. More specifically, knowing some vulnerable function, we can determine whether the firmware of an IoT device contains some security flaw by searching for this function. However, searching for similar vulnerable functions is in general challenging due to the fact that the source code is often not openly available and that it can be compiled for different architectures, using different compilers and compilation settings. In order to handle these varying settings, we can compare the similarity between the graph embeddings derived from the binary functions. In this paper, inspired by the recent advances in deep learning, we propose a new method – GESS (graph embeddings for similarity search) – to derive graph embeddings, and we compare it with various state-of-the-art methods. Our empirical evaluation shows that GESS reaches an AUC of 0.979, thereby outperforming the best known approach. Furthermore, for a fixed low false positive rate, GESS provides a true positive rate (or recall) about 36% higher than the best previous approach. Finally, for a large search space, GESS provides a recall between 50% and 60% higher than the best previous approach.
Mathias Humbert, Didier Dupertuis, Mauro Cherubini, Kévin Huguenin. KGP Meter: Communicating Kin Genomic Privacy to the Masses. In Proceedings of the 7th IEEE European Symposium on Security and Privacy (EuroS&P), 2022.
PDF BibTeX Abstract
```
@INPROCEEDINGS{9797355,
  author={Humbert, Mathias and Dupertuis, Didier and Cherubini, Mauro and Huguenin, Kévin},
  booktitle={2022 IEEE 7th European Symposium on Security and Privacy (EuroS&P)}, 
  title={KGP Meter: Communicating Kin Genomic Privacy to the Masses}, 
  year={2022},
  volume={},
  number={},
  pages={410-429},
  doi={10.1109/EuroSP53844.2022.00033}}

                
```
Direct-to-consumer genetic testing services are gaining momentum: As of today, companies such as 23andMe or AncestryDNA have already attracted 26 million customers. These services raise privacy concerns, exacerbated by the fact that their customers can then share their genomic data on platforms such as GEDmatch. Notwithstanding their right to learn about their genetic background or to share their genomic data, it is paramount that individuals realize that such a behavior damages their relatives' privacy (i.e., kin genomic privacy). In this paper, we present KGP Meter, a new online tool that provides means for raising awareness in the general public about the privacy risks of genomic data sharing. Our tool features various properties that makes it highly interactive, privacy-preserving (i.e., not requiring access to the actual genomic data), and user-friendly. It explores possible configurations in an optimized way and combines well-established graphical models with an entropy-based metric to compute kin genomic privacy scores. Our experiments show that KGP Meter is very responsive. We design and implement an interface that enables users to draw their family trees and indicate which of their relatives' genomes are known, and that communicates the resulting privacy scores to the users. We then analyze the usage of the tool and survey users to better understand users' perceptions towards these risks and evaluate our tool. We observe that most of them find the privacy score worrisome, and that the large majority of them find KGP Meter useful.
Dana Naous, Manus Bonner, Mathias Humbert, Christine Legner. Learning From the Past to Improve the Future : Value-Added Services as a Driver for Mass Adoption of Contact Tracing Apps. Business and Information Systems Engineering, 2022.
PDF BibTeX Abstract
```
@article{naous2022learning,
  title={Learning From the Past to Improve the Future},
  author={Naous, Dana and Bonner, Manus and Humbert, Mathias and Legner, Christine},
  journal={Business \& Information Systems Engineering},
  pages={1--18},
  year={2022},
  publisher={Springer}
}

                
```
Contact tracing apps were considered among the first tools to control the spread of COVID-19 and ease lockdown measures. While these apps can be very effective at stopping transmission and saving lives, the level of adoption remains significantly below the expected critical mass. The public debate as well as academic research about contact tracing apps emphasizes general concerns about privacy (and the associated risks) but often disregards the value-added services, as well as benefits, that can result from a larger user base. To address this gap, the study analyzes goal-congruent features as drivers for user adoption. It uses market research techniques – specifically, conjoint analysis – to study individual and group preferences and gain insights into the prescriptive design. While the results confirm the privacy-preserving design of most European contact tracing apps, they emphasize the role of value-added services in addressing heterogeneous user segments to drive user adoption. The findings thereby are of relevance for designing effective contact tracing apps, but also inform the user-oriented design of apps for health and crisis management that rely on sharing sensitive information.
Lev Velykoivanenko, Kavous Salehzadeh Niksirat, Noé Zufferey, Mathias Humbert, Kévin Huguenin, Mauro Cherubini. Are Those Steps Worth Your Privacy?: Fitness-Tracker Users' Perceptions of Privacy and Utility. Proceedings of the ACM on Interactive, Mobile, Wearable, and Ubiquitous Technologies (IMWUT), vol. 5, issue 4, 2021.
PDF BibTeX Abstract
```
@article{DBLP:journals/imwut/VelykoivanenkoN21,
  author    = {Lev Velykoivanenko and
               Kavous Salehzadeh Niksirat and
               No{\'{e}} Zufferey and
               Mathias Humbert and
               K{\'{e}}vin Huguenin and
               Mauro Cherubini},
  title     = {Are Those Steps Worth Your Privacy?: Fitness-Tracker Users' Perceptions
               of Privacy and Utility},
  journal   = {Proc. {ACM} Interact. Mob. Wearable Ubiquitous Technol.},
  volume    = {5},
  number    = {4},
  pages     = {181:1--181:41},
  year      = {2021},
  url       = {https://doi.org/10.1145/3494960},
  doi       = {10.1145/3494960},
  timestamp = {Thu, 23 Jun 2022 20:04:52 +0200},
  biburl    = {https://dblp.org/rec/journals/imwut/VelykoivanenkoN21.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
Fitness trackers are increasingly popular. The data they collect provides substantial benefits to their users, but it also creates privacy risks. In this work, we investigate how fitness-tracker users perceive the utility of the features they provide and the associated privacy-inference risks. We conduct a longitudinal study composed of a four-month period of fitness-tracker use (N = 227), followed by an online survey (N = 227) and interviews (N = 19). We assess the users' knowledge of concrete privacy threats that fitness-tracker users are exposed to (as demonstrated by previous work), possible privacy-preserving actions users can take, and perceptions of utility of the features provided by the fitness trackers. We study the potential for data minimization and the users' mental models of how the fitness tracking ecosystem works. Our findings show that the participants are aware that some types of information might be inferred from the data collected by the fitness trackers. For instance, the participants correctly guessed that sexual activity could be inferred from heart-rate data. However, the participants did not realize that also the non-physiological information could be inferred from the data. Our findings demonstrate a high potential for data minimization, either by processing data locally or by decreasing the temporal granularity of the data sent to the service provider. Furthermore, we identify the participants' lack of understanding and common misconceptions about how the Fitbit ecosystem works.
Min Chen, Zhikun Zhang, Tianhao Wang, Michael Backes, Mathias Humbert, Yang Zhang. When Machine Unlearning Jeopardizes Privacy. In Proceedings of the 28th ACM Conference on Computer and Communication Security (CCS), 2021.
PDF BibTeX Abstract
```
@inproceedings{DBLP:conf/ccs/Chen000HZ21,
  author    = {Min Chen and
               Zhikun Zhang and
               Tianhao Wang and
               Michael Backes and
               Mathias Humbert and
               Yang Zhang},
  editor    = {Yongdae Kim and
               Jong Kim and
               Giovanni Vigna and
               Elaine Shi},
  title     = {When Machine Unlearning Jeopardizes Privacy},
  booktitle = {{CCS} '21: 2021 {ACM} {SIGSAC} Conference on Computer and Communications
               Security, Virtual Event, Republic of Korea, November 15 - 19, 2021},
  pages     = {896--911},
  publisher = {{ACM}},
  year      = {2021},
  url       = {https://doi.org/10.1145/3460120.3484756},
  doi       = {10.1145/3460120.3484756},
  timestamp = {Tue, 07 Dec 2021 18:36:43 +0100},
  biburl    = {https://dblp.org/rec/conf/ccs/Chen000HZ21.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
The right to be forgotten states that a data owner has the right to erase their data from an entity storing it. In the context of machine learning (ML), the right to be forgotten requires an ML model owner to remove the data owner's data from the training set used to build the ML model, a process known asmachine unlearning. While originally designed to protect the privacy of the data owner, we argue that machine unlearning may leave some imprint of the data in the ML model and thus create unintended privacy risks. In this paper, we perform the first study on investigating the unintended information leakage caused by machine unlearning. We propose a novel membership inference attack that leverages the different outputs of an ML model's two versions to infer whether a target sample is part of the training set of the original model but out of the training set of the corresponding unlearned model. Our experiments demonstrate that the proposed membership inference attack achieves strong performance. More importantly, we show that our attack in multiple cases outperforms the classical membership inference attack on the original ML model, which indicates that machine unlearning can have counterproductive effects on privacy. We notice that the privacy degradation is especially significant for well-generalized ML models where classical membership inference does not perform well. We further investigate four mechanisms to mitigate the newly discovered privacy risks and show that releasing the predicted label only, temperature scaling, and differential privacy are effective. We believe that our results can help improve privacy protection in practical implementations of machine unlearning.
Alexandre Meylan, Mauro Cherubini, Bertil Chapuis, Mathias Humbert, Igor Bilogrevic, Kévin Huguenin. A Study on the Use of Checksums for Integrity Verification of Web Downloads. ACM Transactions on Privacy and Security, vol. 24, issue 1, no 4, 2020.
PDF BibTeX Abstract
```
@article{DBLP:journals/tissec/MeylanCCHBH20,
  author    = {Alexandre Meylan and
               Mauro Cherubini and
               Bertil Chapuis and
               Mathias Humbert and
               Igor Bilogrevic and
               K{\'{e}}vin Huguenin},
  title     = {A Study on the Use of Checksums for Integrity Verification of Web
               Downloads},
  journal   = {{ACM} Trans. Priv. Secur.},
  volume    = {24},
  number    = {1},
  pages     = {4:1--4:36},
  year      = {2020},
  url       = {https://doi.org/10.1145/3410154},
  doi       = {10.1145/3410154},
  timestamp = {Sat, 08 Jan 2022 02:22:55 +0100},
  biburl    = {https://dblp.org/rec/journals/tissec/MeylanCCHBH20.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
App stores provide access to millions of different programs that users can download on their computers. Developers can also make their programs available for download on their websites and host the program files either directly on their website or on third-party platforms, such as mirrors. In the latter case, as users download the software without any vetting from the developers, they should take the necessary precautions to ensure that it is authentic. One way to accomplish this is to check that the published file’s integrity verification code—the checksum—matches that (if provided) of the downloaded file. To date, however, there is little evidence to suggest that such a process is effective. Even worse, very few usability studies about it exist. In this article, we provide the first comprehensive study that assesses the usability and effectiveness of the manual checksum verification process. First, by means of an in-situ experiment with 40 participants and eye-tracking technology, we show that the process is cumbersome and error-prone. Second, after a 4-month-long in-the-wild experiment with 134 participants, we demonstrate how our proposed solution—a Chrome extension that verifies checksums automatically—significantly reduces human errors, improves coverage, and has only limited impact on usability. It also confirms that, sadly, only a tiny minority of websites that link to executable files in our sample provide checksums (0.01%), which is a strong call to action for web standards bodies, service providers, and content creators to increase the use of file integrity verification on their properties.
Inken Hagestedt, Mathias Humbert, Pascal Berrang, Irina Lehmann, Roland Eils, Michael Backes, Yang Zhang. Membership Inference Against DNA Methylation Databases. In Proceedings of the 5th IEEE European Symposium on Security and Privacy (EuroS&P), 2020.
PDF BibTeX Abstract
```
@inproceedings{DBLP:conf/eurosp/HagestedtHBLE0020,
  author    = {Inken Hagestedt and
               Mathias Humbert and
               Pascal Berrang and
               Irina Lehmann and
               Roland Eils and
               Michael Backes and
               Yang Zhang},
  title     = {Membership Inference Against {DNA} Methylation Databases},
  booktitle = {{IEEE} European Symposium on Security and Privacy, EuroS{\&}P
               2020, Genoa, Italy, September 7-11, 2020},
  pages     = {509--520},
  publisher = {{IEEE}},
  year      = {2020},
  url       = {https://doi.org/10.1109/EuroSP48549.2020.00039},
  doi       = {10.1109/EuroSP48549.2020.00039},
  timestamp = {Tue, 10 Nov 2020 11:53:37 +0100},
  biburl    = {https://dblp.org/rec/conf/eurosp/HagestedtHBLE0020.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
Biomedical data sharing is one of the key elements fostering the advancement of biomedical research but poses severe risks towards the privacy of individuals contributing their data, as already demonstrated for genomic data. In this paper, we study whether and to which extent DNA methylation data, one of the most important epigenetic elements regulating human health, is prone to membership inference attacks, a critical type of attack that reveals an individual's participation in a given database. We design and evaluate three different attacks exploiting published summary statistics, among which one is based on machine learning and another is exploiting the dependencies between genome and methylation data. Our extensive evaluation on six datasets containing a diverse set of tissues and diseases collected from more than 1,300 individuals in total shows that such membership inference attacks are effective, even when the target's methylation profile is not accessible. It further shows that the machine-learning approach outperforms the statistical attacks, and that learned models are transferable across different datasets.
Bertil Chapuis, Olamide Omolola, Mauro Cherubini, Mathias Humbert, Kévin Huguenin. An Empirical Study of the Use of Integrity Verification Mechanisms for Web Subresources. In Proceedings of the 2020 World Wide Web Conference (WWW), 2020.
PDF BibTeX Abstract
```
@inproceedings{DBLP:conf/www/ChapuisOCHH20,
  author    = {Bertil Chapuis and
               Olamide Omolola and
               Mauro Cherubini and
               Mathias Humbert and
               K{\'{e}}vin Huguenin},
  editor    = {Yennun Huang and
               Irwin King and
               Tie{-}Yan Liu and
               Maarten van Steen},
  title     = {An Empirical Study of the Use of Integrity Verification Mechanisms
               for Web Subresources},
  booktitle = {{WWW} '20: The Web Conference 2020, Taipei, Taiwan, April 20-24, 2020},
  pages     = {34--45},
  publisher = {{ACM} / {IW3C2}},
  year      = {2020},
  url       = {https://doi.org/10.1145/3366423.3380092},
  doi       = {10.1145/3366423.3380092},
  timestamp = {Mon, 15 Jun 2020 17:03:43 +0200},
  biburl    = {https://dblp.org/rec/conf/www/ChapuisOCHH20.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
Web developers can (and do) include subresources such as scripts, stylesheets and images in their webpages. Such subresources might be stored on content delivery networks (CDNs). This practice creates security and privacy risks, should a subresource be corrupted. The subresource integrity (SRI) recommendation, released in mid-2016 by the W3C, enables developers to include digests in their webpages in order for web browsers to verify the integrity of subresources before loading them. In this paper, we conduct the first large-scale longitudinal study of the use of SRI on the Web by analyzing massive crawls (≈ 3B URLs) of the Web over the last 3.5 years. Our results show that the adoption of SRI is modest (≈), but grows at an increasing rate and is highly influenced by the practices of popular library developers (e.g., Bootstrap) and CDN operators (e.g., jsDelivr). We complement our analysis about SRI with a survey of web developers (N=227): It shows that a substantial proportion of developers know SRI and understand its basic functioning, but most of them ignore important aspects of the recommendation. The results of the survey also show that the integration of SRI by developers is mostly manual – hence not scalable and error prone. This calls for a better integration of SRI in build tools.
Mathias Humbert, Benjamin Trubert, Kévin Huguenin. A Survey on Interdependent Privacy. ACM Computing Surveys, vol. 52, no 6, pp. 1-40, 2020.
PDF BibTeX Abstract
```
@article{DBLP:journals/csur/HumbertTH20,
  author    = {Mathias Humbert and
               Benjamin Trubert and
               K{\'{e}}vin Huguenin},
  title     = {A Survey on Interdependent Privacy},
  journal   = {{ACM} Comput. Surv.},
  volume    = {52},
  number    = {6},
  pages     = {122:1--122:40},
  year      = {2020},
  url       = {https://doi.org/10.1145/3360498},
  doi       = {10.1145/3360498},
  timestamp = {Fri, 27 Mar 2020 08:39:07 +0100},
  biburl    = {https://dblp.org/rec/journals/csur/HumbertTH20.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
The privacy of individuals does not only depend on their own actions and data but may also be affected by the privacy decisions and by the data shared by other individuals. This interdependence is an essential aspect of privacy and ignoring it can lead to serious privacy violations. In this survey, we summarize and analyze research on interdependent privacy risks and on the associated (cooperative and non-cooperative) solutions. We also demonstrate that interdependent privacy has been studied in isolation in different research communities. By doing so, we systematize knowledge on interdependent privacy research and provide insights on how this research should be conducted and which challenges it should address.
Yang Zhang, Mathias Humbert, Bartlomiej Surma, Praveen Manoharan, Jilles Vreeken, Michael Backes. Towards Plausible Graph Anonymization. In Proceedings of the 27th Annual Network and Distributed System Security Symposium (NDSS), 2020.
PDF BibTeX Abstract
```
@inproceedings{DBLP:conf/ndss/0016HSMV020,
  author    = {Yang Zhang and
               Mathias Humbert and
               Bartlomiej Surma and
               Praveen Manoharan and
               Jilles Vreeken and
               Michael Backes},
  title     = {Towards Plausible Graph Anonymization},
  booktitle = {27th Annual Network and Distributed System Security Symposium, {NDSS}
               2020, San Diego, California, USA, February 23-26, 2020},
  publisher = {The Internet Society},
  year      = {2020},
  url       = {https://www.ndss-symposium.org/ndss-paper/towards-plausible-graph-anonymization/},
  timestamp = {Tue, 21 Jul 2020 16:10:07 +0200},
  biburl    = {https://dblp.org/rec/conf/ndss/0016HSMV020.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
Social graphs derived from online social interactions contain a wealth of information that is nowadays extensively used by both industry and academia. However, as social graphs contain sensitive information, they need to be properly anonymized before release. Most of the existing graph anonymization mechanisms rely on the perturbation of the original graph’s edge set. In this paper, we identify a fundamental weakness of these mechanisms: They neglect the strong structural proximity between friends in social graphs, thus add implausible fake edges for anonymization. To exploit this weakness, we first propose a metric to quantify an edge’s plausibility by relying on graph embedding. Extensive experiments on three real-life social network datasets demonstrate that our plausibility metric can very effectively differentiate fake edges from original edges with AUC (area under the ROC curve) values above 0.95 in most of the cases. We then rely on a Gaussian mixture model to automatically derive the threshold on the edge plausibility values to determine whether an edge is fake, which enables us to recover to a large extent the original graph from the anonymized graph. We further demonstrate that our graph recovery attack jeopardizes the privacy guarantees provided by the considered graph anonymization mechanisms. To mitigate this vulnerability, we propose a method to generate fake yet plausible edges given the graph structure and incorporate it into the existing anonymization mechanisms. Our evaluation demonstrates that the enhanced mechanisms decrease the chances of graph recovery, reduce the success of graph de- anonymization (up to 30%), and provide even better utility than the existing anonymization mechanisms.
Spyros Boukoros, Mathias Humbert, Stefan Katzenbeisser, Carmela Troncoso. On (The Lack Of) Location Privacy in Crowdsourcing Applications. In Proceedings of the 28th USENIX Security Symposium, 2019.
PDF BibTeX Abstract Code
```
@inproceedings{boukoros2019lack,
  title={On (the lack of) location privacy in crowdsourcing applications},
  author={Boukoros, Spyros and Humbert, Mathias and Katzenbeisser, Stefan and Troncoso, Carmela},
  booktitle={28th $\{$USENIX$\}$ Security Symposium ($\{$USENIX$\}$ Security 19)},
  pages={1859--1876},
  year={2019}
}

                
```
Crowdsourcing enables application developers to benefit from large and diverse datasets at a low cost. Specifically, mobile crowdsourcing (MCS) leverages users’ devices as sensors to perform geo-located data collection. The collection of geolocated data raises serious privacy concerns for users. Yet, despite the large research body on location privacy-preserving mechanisms (LPPMs), MCS developers implement little to no protection for data collection or publication. To understand this mismatch, we study the performance of existing LPPMs on publicly available data from two mobile crowdsourcing projects. Our results show that well-established defenses are either not applicable or offer little protection in the MCS setting. Additionally, they have a much stronger impact on applications’ utility than foreseen in the literature. This is because existing LPPMs, designed with location-based services (LBSs) in mind, are optimized for utility functions based on users’ locations, while MCS utility functions depend on the values (e.g., measurements) associated with those locations. We finally outline possible research avenues to facilitate the development of new location privacy solutions that fit the needs of MCS so that the increasing number of such applica- tions do not jeopardize their users’ privacy.
Inken Hagestedt, Yang Zhang, Mathias Humbert, Pascal Berrang, Haixu Tang, Xiaofeng Wang, Michael Backes. MBeacon: Privacy-Preserving Beacons for DNA Methylation Data. In Proceedings of the 26th Annual Network and Distributed System Security Symposium (NDSS), 2019. Distinguished Paper Award
PDF BibTeX Abstract
```
@inproceedings{DBLP:conf/ndss/Hagestedt0HBT0019,
  author    = {Inken Hagestedt and
               Yang Zhang and
               Mathias Humbert and
               Pascal Berrang and
               Haixu Tang and
               XiaoFeng Wang and
               Michael Backes},
  title     = {MBeacon: Privacy-Preserving Beacons for {DNA} Methylation Data},
  booktitle = {26th Annual Network and Distributed System Security Symposium, {NDSS}
               2019, San Diego, California, USA, February 24-27, 2019},
  year      = {2019},
  crossref  = {DBLP:conf/ndss/2019},
  url       = {https://www.ndss-symposium.org/ndss-paper/mbeacon-privacy-preserving-beacons-for-dna-methylation-data/},
  timestamp = {Thu, 02 May 2019 15:52:50 +0200},
  biburl    = {https://dblp.org/rec/bib/conf/ndss/Hagestedt0HBT0019},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

                
```
The advancement of molecular profiling techniques fuels biomedical research with a deluge of data. To facilitate data sharing, the Global Alliance for Genomics and Health established the Beacon system, a search engine designed to help researchers find datasets of interest. While the current Beacon system only supports genomic data, other types of biomedical data, such as DNA methylation, are also essential for advancing our understanding in the field. In this paper, we propose the first Beacon system for DNA methylation data sharing: MBeacon. As the current genomic Beacon is vulnerable to privacy attacks, such as membership inference, and DNA methylation data is highly sensitive, we take a privacy-by-design approach to construct MBeacon.
First, we demonstrate the privacy threat, by proposing a membership inference attack tailored specifically to unprotected methylation Beacons. Our experimental results show that 100 queries are sufficient to achieve a successful attack with AUC (area under the ROC curve) above 0.9. To remedy this situation, we propose a novel differential privacy mechanism, namely SVT^2, which is the core component of MBeacon. Extensive experiments over multiple datasets show that SVT^2 can successfully mitigate membership privacy risks without significantly harming utility. We further implement a fully functional prototype of MBeacon which we make available to the research community.
Ahmed Salem, Yang Zhang, Mathias Humbert, Pascal Berrang, Mario Fritz, Michael Backes. ML-Leaks: Model and Data Independent Membership Inference Attacks and Defenses on Machine Learning Models. In Proceedings of the 26th Annual Network and Distributed System Security Symposium (NDSS), 2019.
PDF BibTeX Abstract Code
```
@inproceedings{DBLP:conf/ndss/Salem0HBF019,
  author    = {Ahmed Salem and
               Yang Zhang and
               Mathias Humbert and
               Pascal Berrang and
               Mario Fritz and
               Michael Backes},
  title     = {ML-Leaks: Model and Data Independent Membership Inference Attacks
               and Defenses on Machine Learning Models},
  booktitle = {26th Annual Network and Distributed System Security Symposium, {NDSS}
               2019, San Diego, California, USA, February 24-27, 2019},
  year      = {2019},
  crossref  = {DBLP:conf/ndss/2019},
  url       = {https://www.ndss-symposium.org/ndss-paper/ml-leaks-model-and-data-independent-membership-inference-attacks-and-defenses-on-machine-learning-models/},
  timestamp = {Thu, 02 May 2019 15:52:50 +0200},
  biburl    = {https://dblp.org/rec/bib/conf/ndss/Salem0HBF019},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}
                
```
Machine learning (ML) has become a core component of many real-world applications and training data is a key factor that drives current progress. This huge success has led Internet companies to deploy machine learning as a service (MLaaS). Recently, the first membership inference attack has shown that extraction of information on the training set is possible in such MLaaS settings, which has severe security and privacy implications.
However, the early demonstrations of the feasibility of such attacks have many assumptions on the adversary, such as using multiple so-called shadow models, knowledge of the target model structure, and having a dataset from the same distribution as the target model’s training data. We relax all these key assumptions, thereby showing that such attacks are very broadly applicable at low cost and thereby pose a more severe risk than previously thought. We present the most comprehensive study so far on this emerging and developing threat using eight diverse datasets which show the viability of the proposed attacks across domains. In addition, we propose the first effective defense mechanisms against such broader class of membership inference attacks that maintain a high level of utility of the ML model.
Ahmed Salem, Pascal Berrang, Mathias Humbert, Michael Backes. Privacy-Preserving Similar Patient Queries for Combined Biomedical Data. Proceedings on Privacy Enhancing Technologies (PoPETS), 2019
PDF BibTeX Abstract
```
@article{DBLP:journals/popets/SalemBHB19,
  author    = {Ahmed Salem and
               Pascal Berrang and
               Mathias Humbert and
               Michael Backes},
  title     = {Privacy-Preserving Similar Patient Queries for Combined Biomedical
               Data},
  journal   = {PoPETs},
  volume    = {2019},
  number    = {1},
  pages     = {47--67},
  year      = {2019},
  url       = {https://doi.org/10.2478/popets-2019-0004},
  doi       = {10.2478/popets-2019-0004},
  timestamp = {Tue, 26 Feb 2019 17:22:18 +0100},
  biburl    = {https://dblp.org/rec/bib/journals/popets/SalemBHB19},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}
                
```
The decreasing costs of molecular profiling have fueled the biomedical research community with a plethora of new types of biomedical data, enabling a breakthrough towards more precise and personalized medicine. Naturally, the increasing availability of data also enables physicians to compare patients’ data and treatments easily and to find similar patients in order to propose the optimal therapy. Such similar patient queries (SPQs) are of utmost importance to medical practice and will be relied upon in future health information exchange systems. While privacy-preserving solutions have been previously studied, those are limited to genomic data, ignoring the different newly available types of biomedical data.

In this paper, we propose new cryptographic techniques for finding similar patients in a privacy-preserving manner with various types of biomedical data, including genomic, epigenomic and transcriptomic data as well as their combination. We design protocols for two of the most common similarity metrics in biomedicine: the Euclidean distance and Pearson correlation coefficient. Moreover, unlike previous approaches, we account for the fact that certain locations contribute differently to a given disease or phenotype by allowing to limit the query to the relevant locations and to assign them different weights. Our protocols are specifically designed to be highly efficient in terms of communication and bandwidth, requiring only one or two rounds of communication and thus enabling scalable parallel queries. We rigorously prove our protocols to be secure based on cryptographic games and instantiate our technique with three of the most important types of biomedical data – namely DNA, microRNA expression, and DNA methylation. Our experimental results show that our protocols can compute a similarity query over a typical number of positions against a database of 1,000 patients in a few seconds. Finally, we propose and formalize strategies to mitigate the threat of malicious users or hospitals.
Alexandra Olteanu, Mathias Humbert, Kévin Huguenin, Jean-Pierre Hubaux. The (Co-)Location Sharing Game. Proceedings on Privacy Enhancing Technologies (PoPETS), 2019.
PDF BibTeX Abstract
```
@article{
}
                
```
Most popular location-based social networks, such as Facebook and Foursquare, let their (mobile) users post location and co-location (involving other users) information. Such posts bring social benefits to the users who post them but also to their friends who view them. Yet, they also represent a severe threat to the users’ privacy, as co-location information introduces interdependences between users. We propose the first game-theoretic framework for analyzing the strategic behaviors, in terms of information sharing, of users of OSNs. To design parametric utility functions that are representative of the users’ actual preferences, we also conduct a survey of 250 Facebook users and use con-joint analysis to quantify the users’ benefits of sharing vs. viewing (co)-location information and their preference for privacy vs. benefits. Our survey findings expose the fact that, among the users, there is a large variation, in terms of these preferences. We extensively evaluate our framework through data-driven numerical simulations. We study how users’ individual preferences influence each other’s decisions, we identify several factors that significantly affect these decisions (among which, the mobility data of the users), and we determine situations where dangerous patterns can emerge (e.g., a vicious circle of sharing, or an incentive to over-share) – even when the users share similar preferences.
Mauro Cherubini, Alexandre Meylan, Bertil Chapuis, Mathias Humbert, Igor Bilogrevic, Kévin Huguenin. Towards Usable Checksums: Automating the Integrity Verification of Web Downloads for the Masses. In Proceedings of the 25th ACM Conference on Computer and Communications Security (CCS), 2018.
PDF BibTeX Abstract Code Demo
```
@inproceedings{cherubini2018towards,
  title={Towards Usable Checksums: Automating the Integrity Verification of Web Downloads for the Masses},
  author={Cherubini, Mauro and Meylan, Alexandre and Chapuis, Bertil and Humbert, Mathias and Bilogrevic, Igor and Huguenin, K{\'e}vin},
  booktitle={ACM CCS}
  year={2018}
}
                
```
Internet users can download software for their computers from app stores (e.g., Mac App Store and Windows Store) or from other sources, such as the developers' websites. Most Internet users in the US rely on the latter, according to our representative study, which makes them directly responsible for the content they download. To enable users to detect if the downloaded files have been corrupted, developers can publish a checksum together with the link to the program file; users can then manually verify that the checksum matches the one they obtain from the downloaded file. In this paper, we assess the prevalence of such behavior among the general Internet population in the US (N=2,000), and we develop easy-to-use tools for users and developers to automate both the process of checksum verification and generation. Specifically, we propose an extension to the recent W3C specification for sub-resource integrity in order to provide integrity protection for download links. Also, we develop an extension for the popular Chrome browser that computes and verifies checksums of downloaded files automatically, and an extension for the WordPress CMS that developers can use to easily attach checksums to their remote content. Our in situ experiments with 40 participants demonstrate the usability and effectiveness issues of checksums verification, and shows user desirability for our extension.
Yang Zhang, Mathias Humbert, Tahleen Rahman, Cheng-Te Li, Jun Pang, Michael Backes. Tagvisor: A Privacy Advisor for Sharing Hashtags. In Proceedings of the 2018 World Wide Web Conference (WWW), 2018.
PDF BibTeX Abstract
```
@inproceedings{zhang2018tagvisor,
  title={Tagvisor: A privacy advisor for sharing hashtags},
  author={Zhang, Yang and Humbert, Mathias and Rahman, Tahleen and Li, Cheng-Te and Pang, Jun and Backes, Michael},
  booktitle={Proceedings of The Web Conference 2018 (WWW'18)},
  pages={287--296},
  year={2018},
  organization={ACM Press}
}
                
```
Hashtags have emerged as a widely used concept of popular culture and campaigns, but their implications on people’s privacy have not been investigated so far. In this paper, we present the first systematic analysis of privacy issues induced by hashtags. We concentrate in particular on location, which is recognized as one of the key privacy concerns in the Internet era. By relying on a random forest model, we show that we can infer a user’s precise location from hashtags with accuracy of 70% to 76%, depending on the city. To remedy this situation, we introduce a system called Tagvisor that systematically suggests alternative hashtags if the user-selected ones constitute a threat to location privacy. Tagvisor realizes this by means of three conceptually different obfuscation techniques and a semantics-based metric for measuring the consequent utility loss. Our findings show that obfuscating as little as two hashtags already provides a near-optimal trade-off between privacy and utility in our dataset. This in particular renders Tagvisor highly time-efficient, and thus, practical in real-world settings.
Pascal Berrang, Mathias Humbert, Yang Zhang, Irina Lehmann, Roland Eils, Michael Backes. Dissecting Privacy Risks in Biomedical Data. In Proceedings of the 3rd IEEE European Symposium on Security and Privacy (EuroS&P), 2018.
PDF BibTeX Abstract
```
@inproceedings{BHZL+18,
 author	  =  	{Berrang, Pascal and Humbert, Mathias and Zhang, Yang and Lehmann, Irina and Eils, Roland and Backes, Michael},
 title	  =  	{Dissecting Privacy Risks in Biomedical Data},
 booktitle	  =  	{IEEE EuroS\&P},
 year	  =  	{2018}
}
                
```
The decreasing costs of molecular profiling has fueled the biomedical research community with a plethora of new types of biomedical data, enabling a breakthrough towards a more precise and personalized medicine. However, the release of these intrinsically highly sensitive data poses a new severe privacy threat. While biomedical data is largely associated with our health, there also exist various correlations between different types of biomedical data, along the temporal dimension, and also in-between family members. However, so far, the security community has focused on privacy risks stemming from genomic data, largely overlooking the manifold interdependencies between other biomedical data.

In this paper, we present a generic framework for quantifying the privacy risks in biomedical data taking into account the various interdependencies between data (i) of different types, (ii) from different individuals, and (iii) at different time. To this end, we rely on a Bayesian network model that allows us to take all aforementioned dependencies into account and run exact probabilistic inference attacks very efficiently. Furthermore, we introduce a generic algorithm for building the Bayesian network, which encompasses expert knowledge for known dependencies, such as genetic inheritance laws, and learns previously unknown dependencies from the data. Then, we conduct a thorough inference risk evaluation with a very rich dataset containing genomic and epigenomic data of mothers and children over multiple years. Besides effective probabilistic inference, we further demonstrate that our Bayesian network model can also serve as a building block for other attacks. We show that, with our framework, an adversary can efficiently identify the parent-child relationships based on methylation data with a success rate of 95%.
Erman Ayday, Mathias Humbert. Inference Attacks against Kin Genomic Privacy. IEEE Security & Privacy, vol. 15, no 5, pp. 29-37, 2017.
PDF BibTeX Abstract
```
@article{ayday2017inference,
  title={Inference attacks against kin genomic privacy},
  author={Ayday, Erman and Humbert, Mathias},
  journal={IEEE Security \& Privacy},
  number={5},
  pages={29--37},
  year={2017},
  publisher={IEEE}
}
                
```
Genomic data poses serious interdependent risks: your data might also leak information about your family members' data. Methods attackers use to infer genomic information, as well as recent proposals for enhancing genomic privacy, are discussed.
¹Michael Backes, Mathias Humbert, Jun Pang, Yang Zhang. walk2friends: Inferring Social Links from Mobility Profiles. In Proceedings of the 24th ACM SIGSAC Conference on Computer and Communications Security (CCS), 2017.
PDF BibTeX Abstract Code
```
@inproceedings{BHPZ17,
 author	  =  	{Backes, Michael and and Humbert, Mathias and Pang, Jun and Zhang, Yang},
 title	  =  	{walk2friends: Inferring Social Links from Mobility Profiles},
 booktitle	  =  	{ACM CCS},
 year	  =  	{2017}
}
                
```
The development of positioning technologies has resulted in an increasing amount of mobility data being available. While bringing a lot of convenience to people’s life, such availability also raises serious concerns about privacy. In this paper, we concentrate on one of the most sensitive information that can be inferred from mobility data, namely social relationships. We propose a novel social relation inference attack that relies on an advanced feature learning technique to automatically summarize users’ mobility features. Compared to existing approaches, our attack is able to predict any two individuals’ social relation, and it does not require the adversary to have any prior knowledge on existing social relations. These advantages signi cantly increase the applicability of our attack and the scope of the privacy assessment. Extensive experiments conducted on a large dataset demonstrate that our inference attack is effective, and achieves between 13% to 20% improvement over the best state-of-the-art scheme. We propose three defense mechanisms – hiding, replacement and generalization – and evaluate their effectiveness for mitigating the social link privacy risks stemming from mobility data sharing. Our experimental results show that both hiding and replacement mechanisms outperform generalization. Moreover, hiding and replacement achieve a comparable trade-off between utility and privacy, the former preserving better utility and the latter providing better privacy.
¹Michael Backes, Pascal Berrang, Matthias Bieg, Roland Eils, Carl Herrmann, Mathias Humbert, Irina Lehmann. Identifying Personal DNA Methylation Profiles by Genotype Inference. In Proceedings of the 38th IEEE Symposium on Security and Privacy (S&P), 2017.
PDF BibTeX Abstract Code
```
@inproceedings{BBBE+17,
 author	  =  	{Backes, Michael and Berrang, Pascal and Bieg, Matthias and Eils, Roland and Herrmann, Carl and Humbert, Mathias and Lehmann, Irina},
 title	  =  	{Identifying Personal DNA Methylation Profiles by Genotype Inference},
 booktitle	  =  	{IEEE S\&P},
 year	  =  	{2017}
}
                
```
Since the first whole-genome sequencing, the biomedical research community has made significant steps towards a more precise, predictive and personalized medicine. Genomic data is nowadays widely considered privacy-sensitive and consequently protected by strict regulations and released only after careful consideration. Various additional types of biomedical data, however, are not shielded by any dedicated legal means and consequently disseminated much less thoughtfully. This in particular holds true for DNA methylation data as one of the most important and well-understood epigenetic element influencing human health.

In this paper, we show that, in contrast to the aforementioned belief, releasing one's DNA methylation data causes privacy issues akin to releasing one's actual genome. We show that already a small subset of methylation regions influenced by genomic variants are sufficient to infer parts of someone's genome, and to further map this DNA methylation profile to the corresponding genome. Notably, we show that such re-identification is possible with 97.5\% accuracy, relying on a dataset of more than 2500 genomes, and that we can reject all wrongly matched genomes using an appropriate statistical test. We provide means for countering this threat by proposing a novel cryptographic scheme for privately classifying tumors that enables a privacy-respecting medical diagnosis in a common clinical setting. The scheme relies on a combination of random forests and homomorphic encryption, and it is proven secure in the honest-but-curious model. We evaluate this scheme on real DNA methylation data, and show that we can keep the computational overhead to acceptable values for our application scenario.
Florian Tramèr, Vaggelis Atlidakis, Roxana Geambasu, Daniel Hsu, Jean-Pierre Hubaux, Mathias Humbert, Ari Juels, Huang Lin. FairTest: Discovering Unwarranted Associations in Data-Driven Applications. In Proceedings of the 2nd IEEE European Symposium on Security and Privacy (EuroS&P), 2017.
PDF BibTeX Abstract Code
```
@inproceedings{TAGH+17,
 author	  =  	{Tramèr, Florian and Atlidakis, Vaggelis and Geambasu, Roxana and Hsu, Daniel and Hubaux, Jean-Pierre and Humbert, Mathias and Juels, Ari and Lin, Huang},
 title	  =  	{FairTest: Discovering Unwarranted Associations in Data-Driven Applications},
 booktitle	  =  	{IEEE EuroS\&P},
 year	  =  	{2017}
}
                
```
Abstract—In a world where traditional notions of privacy are increasingly challenged by the myriad companies that collect and analyze our data, it is important that decision-making entities are held accountable for unfair treatments arising from irresponsible data usage. Unfortunately, a lack of appropriate methodologies and tools means that even identifying unfair or discriminatory effects can be a challenge in practice. We introduce the unwarranted associations (UA) framework, a principled methodology for the discovery of unfair, discriminatory, or offensive user treatment in data-driven applications. The UA framework unifies and rationalizes a number of prior attempts at formalizing algorithmic fairness. It uniquely combines multiple investigative primitives and fairness metrics with broad applicability, granular exploration of unfair treatment in user subgroups, and incorporation of natural notions of utility that may account for observed disparities. We instantiate the UA framework in FairTest, the first comprehensive tool that helps developers check data-driven applications for unfair user treatment. It enables scalable and statistically rigorous investigation of associations between application outcomes (such as prices or premiums) and sensitive user attributes (such as race or gender). Furthermore, FairTest provides debugging capabilities that let programmers rule out potential confounders for observed unfair effects. We report on use of FairTest to investigate and in some cases address disparate impact, offensive labeling, and uneven rates of algorithmic error in four data-driven applications. As examples, our results reveal subtle biases against older populations in the distribution of error in a predictive health application and offensive racial labeling in an image tagger.
Mathias Humbert, Erman Ayday, Jean-Pierre Hubaux, Amalio Telenti. Quantifying Interdependent Risks in Genomic Privacy. ACM Transactions on Privacy and Security (TOPS), vol. 20, no 1, pp. 1-31, 2017.
PDF BibTeX Abstract
```
@inproceedings{humbert2017,
author={M. Humbert and E. Ayday and J.-P. Hubaux and A. Telenti}, 
journal={ACM Transactions on Privacy and Security}, 
title={Quantifying Interdependent Risks in Genomic Privacy}, 
year={2017}, 
volume={20}, 
number={1}, 
pages={1-31}, 
ISSN={2471-2566}, 
month={February},
}
                
```
The rapid progress in human-genome sequencing is leading to a high availability of genomic data. These data are notoriously very sensitive and stable in time, and highly correlated among relatives. In this paper, we study the implications of these familial correlations on kin genomic privacy. We formalize the problem and detail efficient reconstruction attacks based on graphical models and belief propagation. With our approach, an attacker can infer the genomes of the relatives of an individual whose genome or phenotype are observed, by notably relying on Mendel’s Laws, statistical relationships between the genomic variants, and between the genome and the phenotype. We evaluate the effect of these dependencies on privacy with respect to the amount of observed variants and the relatives sharing them. We also study how the algorithmic performance evolves when we take these various relationships into account. Furthermore, to quantify the level of genomic privacy as a result of the proposed inference attack, we discuss possible definitions of genomic privacy metrics, and compare their values and evolution. Genomic data reveals Mendelian disorders and the likelihood of developing severe diseases such as Alzheimer’s. We also introduce the quantification of health privacy, specifically the measure of how well the predisposition to a disease is concealed from an attacker. We evaluate our approach on actual genomic data from a pedigree and show the threat extent by combining data gathered from a genome-sharing website and from an online social network.
Alexandra-Mihaela Olteanu, Kévin Huguenin, Reza Shokri, Mathias Humbert, Jean-Pierre Hubaux. Quantifying Interdependent Privacy Risks with Location Data. IEEE Transactions on Mobile Computing (TMC), vol. 16, no 3, pp. 829-842, 2017.
PDF BibTeX Abstract
```
@inproceedings{olteanu2016colocation,
author={A. M. Olteanu and K. Huguenin and R. Shokri and M. Humbert and J. P. Hubaux}, 
journal={IEEE Transactions on Mobile Computing}, 
title={Quantifying Interdependent Privacy Risks with Location Data}, 
year={2017}, 
volume={16}, 
number={3}, 
pages={829-842}, 
ISSN={1536-1233}, 
month={March},
}
                
```
Co-location information about users is increasingly available online. For instance, mobile users more and more frequently report their co-locations with other users in the messages and in the pictures they post on social networking websites by tagging the names of the friends they are with. The users’ IP addresses also constitute a source of co-location information. Combined with (possibly obfuscated) location information, such co-locations can be used to improve the inference of the users’ locations, thus further threatening their location privacy: As co-location information is taken into account, not only a user’s reported locations and mobility patterns can be used to localize her, but also those of her friends (and the friends of their friends and so on). In this paper, we study this problem by quantifying the effect of co-location information on location privacy, considering an adversary such as a social network operator that has access to such information. We formalize the problem and derive an optimal inference algorithm that incorporates such co-location information, yet at the cost of high complexity. We propose some approximate inference algorithms, including a solution that relies on the belief propagation algorithm executed on a general Bayesian network model, and we extensively evaluate their performance. Our experimental results show that, even in the case where the adversary considers co-locations of the targeted user with a single friend, the median location privacy of the user is decreased by up to 62% in a typical setting. We also study the effect of the different parameters (e.g., the settings of the location-privacy protection mechanisms) in different scenarios.
¹Michael Backes, Pascal Berrang, Mathias Humbert, Xiaoyu Shen, Verena Wolf. Simulating the Large-scale Erosion of Genomic Privacy Over Time. 3rd International Workshop on Genome Privacy and Security (GenoPri), 2016. Selected for publication in IEEE/ACM Transactions on Computational Biology and Bioinformatics.
PDF BibTeX Abstract
```
@inproceedings{backes2016erosion,
   author      = {Backes, Michael and Berrang, Pascal and Humbert, Mathias and Shen, Xiaoyu and Wolf, Verena},
   booktitle   = {3rd International Workshop on Genome Privacy and Security (GenoPri)},
   title       = {Simulating the Large-scale Erosion of Genomic Privacy Over Time},
   year={2016}
}
                
```
The dramatically decreasing costs of DNA sequencing have triggered more than a million humans to date to have their genotypes sequenced. Moreover, these individuals increasingly make their genomic data publicly available, and thereby create unique privacy threats not only for themselves, but also for their relatives because of their DNA similarities. More generally, an entity that gains access to a significant fraction of sequenced genotypes from a given population might be able to infer even the genomes of unsequenced individuals by relying on available data.

In this paper, we propose a simulation-based model for quantifying the impact of continuously sequencing and publicizing personal genomic data on a population’s genomic privacy. Our simulation probabilistically models data sharing by individuals and additionally takes into account the influence on genomic privacy of geopolitical events such as migration, and sociological trends such as interracial marriage. We exemplarily instantiate our simulation with a sample population of 1,000 individuals, and evaluate the evolution of privacy under different settings over either thousands of genomic variants or a subset of variants influencing the phenotype. Our findings notably demonstrate that an increasing sharing rate of genomic data in the future entails a substantial negative effect on the privacy of all older generations. Moreover, we find that mixed populations, due to their large genomic diversity, face a less severe erosion of genomic privacy over time than more homogeneous populations. However, even when no data is shared, the genomic privacy averaged over a large number of variants is already very low since mere population allele frequencies already reveal a lot of information about the values of the genomic variants. By focusing on a subset of sensitive variants, we observe a higher genetic diversity in the population. Thus, genomic-data sharing can be much more detrimental for the privacy of the most sensitive variants.
¹Michael Backes, Pascal Berrang, Mathias Humbert, Praveen Manoharan. Membership Privacy in MicroRNA-based Studies. In Proceedings of the 23rd ACM SIGSAC Conference on Computer and Communications Security (CCS), 2016.
PDF BibTeX Abstract
```
@inproceedings{backes2016membership,
   author      = {Backes, Michael and Berrang, Pascal and Humbert, Mathias and Manoharan, Praveen},
   booktitle   = {Proceedings of the 23rd ACM Conference on Computer and Communications Security (CCS)},
   title       = {{M}embership {P}rivacy in {M}icro{RNA}-based {S}tudies},
   year={2016}
}
                
```
The continuous decrease in cost of molecular profiling tests is revolutionizing medical research and practice, but it also raises new privacy concerns. One of the first attacks against privacy of biological data, proposed by Homer et al. in 2008, showed that, by knowing parts of the genome of a given individual and summary statistics of a genome-based study, it is possible to detect if this individual participated in the study. Since then, a lot of work has been carried out to further study the theoretical limits and to counter the genome-based membership inference attack. However, genomic data are by no means the only or the most influential biological data threatening personal privacy. For instance, whereas the genome informs us about the risk of developing some diseases in the future, epigenetic biomarkers, such as microRNAs, are directly and deterministically affected by our health condition including most common severe diseases.

In this paper, we show that the membership inference attack also threatens the privacy of individuals contributing their microRNA expressions to scientific studies. Our results on real and public microRNA expression data demonstrate that disease-specific datasets are especially prone to membership detection, offering a true-positive rate of up to 77% at a false-negative rate of less than 1%. We present two attacks: one relying on the L1 distance and the other based on the likelihood-ratio test. We show that the likelihood-ratio test provides the highest adversarial success and we derive a theoretical limit on this success. In order to mitigate the membership inference, we propose and evaluate both a differentially private mechanism and a hiding mechanism. We also consider two types of adversarial prior knowledge for the differentially private mechanism and show that, for relatively large datasets, this mechanism can protect the privacy of participants in miRNA-based studies against strong adversaries without degrading the data utility too much. Based on our findings and given the current number of miRNAs, we recommend to only release summary statistics of datasets containing at least a couple of hundred individuals.
¹Michael Backes, Pascal Berrang, Anne Hecksteden, Mathias Humbert, Andreas Keller, Tim Meyer. Privacy in Epigenetics: Temporal Linkability of MicroRNA Expression Profiles. In Proceedings of the 25th USENIX Security Symposium, 2016.
PDF BibTeX Abstract
```
@inproceedings{backes2016epigenetics,
   author      = {Backes, Michael and Berrang, Pascal and Hecksteden, Anne and Humbert, Mathias and Keller, Andreas and Meyer, Tim},
   booktitle   = {Proceedings of the 25th USENIX Security Symposium},
   title       = {Privacy in {E}pigenetics: {T}emporal {L}inkability of {M}icroRNA {E}xpression {P}rofiles},
	year={2016}
}
                
```
The decreasing cost of molecular profiling tests, such as DNA sequencing, and the consequent increasing availability of biological data are revolutionizing medicine, but at the same time create novel privacy risks. The research community has already proposed a plethora of methods for protecting genomic data against these risks. However, the privacy risks stemming from epigenetics, which bridges the gap between the genome and our health characteristics, have been largely overlooked so far, even though epigenetic data such as microRNAs (miRNAs) are no less privacy sensitive. This lack of investigation is attributed to the common belief that the inherent temporal variability of miRNAs shields them from being tracked and linked over time.

In this paper, we show that, contrary to this belief, miRNA expression profiles can be successfully tracked over time, despite their variability. Specifically, we show that two blood-based miRNA expression profiles taken with a time difference of one week from the same person can be matched with a success rate of 90%. We furthermore observe that this success rate stays almost constant when the time difference is increased from one week to one year. In order to mitigate the linkability threat, we propose and thoroughly evaluate two countermeasures: (i) hiding a subset of disease-irrelevant miRNA expressions, and (ii) probabilistically sanitizing the miRNA expression profiles. Our experiments show that the second mechanism provides a better trade-off between privacy and disease-prediction accuracy.

¹ Authors listed in alphabetical order.
2010 - 2015

Mathias Humbert, Kévin Huguenin, Joachim Hugonot, Erman Ayday, Jean-Pierre Hubaux. De-anonymizing Genomic Databases with Phenotypic Traits. Proceedings on Privacy Enhancing Technologies (PoPETS), 2015.
PDF BibTeX Abstract
```
@inproceedings{humbert2015non,
   author      = {Humbert, Mathias and Huguenin, K{\'e}vin and Hugonot, Joachim and Ayday, Erman and Hubaux, Jean-Pierre},
   booktitle   = {Proceedings on Privacy Enhancing Technologies},
   title       = {{D}e-anonymizing {G}enomic {D}atabases with {P}henotypic {T}raits},
	year={2015},
	publisher={De Gruyter}
}
                
```
People increasingly have their genomes sequenced and some of them share their genomic data online. They do so for various purposes, including to find relatives and to help advance genomic research. An individual's genome carries very sensitive, private information such as its owner's susceptibility to diseases, which could be used for discrimination. Therefore, genomic databases are often anonymized. However, an individual's genotype is also linked to visible phenotypic traits, such as eye or hair color, which can be used to re-identify users in anonymized public genomic databases, thus raising severe privacy issues. For instance, an adversary can identify a target's genome using known her phenotypic traits and subsequently infer her susceptibility to Alzheimer's disease.

In this paper, we quantify, based on various phenotypic traits, the extent of this threat in several scenarios by implementing de-anonymization attacks on a genomic database of OpenSNP users sequenced by 23andMe. Our experimental results show that the proportion of correct matches reaches 23% with a supervised approach in a database of 50 participants. Our approach outperforms the baseline by a factor of four, in terms of the proportion of correct matches, in most scenarios. We also evaluate the adversary's ability to predict individuals' predisposition to Alzheimer's disease, and we observe that the inference error can be halved compared to the baseline. We also analyze the effect of the number of known phenotypic traits on the success rate of the attack. As progress is made in genomic research, especially for genotype-phenotype associations, the threat presented in this paper will become more serious.
Mathias Humbert. When Others Impinge upon Your Privacy: Interdependent Risks and Protection in a Connected World. Ph.D. Thesis, Ecole Polytechnique Fédérale de Lausanne (EPFL), 2015.
Advisor: Prof. Jean-Pierre Hubaux
Jury: Prof. Jacques Fellay, Prof. Ari Juels, Prof. Refik Molva, Prof. Emre Telatar
PDF BibTeX Abstract
```
@phdthesis{humbert2015others,
  title={When {O}thers {I}mpinge upon {Y}our {P}rivacy: {I}nterdependent {R}isks and {P}rotection in a {C}onnected {W}orld},
  author={Humbert, Mathias},
  year={2015},
  school={{\'E}cole {P}olytechnique {F}{\'e}D{\'e}rale de {L}anne}
}
                
```
Privacy is defined as the right to control, edit, manage, and delete information about oneself and decide when, how, and to what extent this information is communicated to others. Therefore, every person should ideally be empowered to manage and protect his own data, individually and independently of others. This assumption, however, barely holds in practice, because people are by nature biologically and socially interconnected. An individual's identity is essentially determined at the biological and social levels. First, a person is biologically determined by his DNA, his genes, that fully encode his physical characteristics. Second, human beings are social animals, with a strong need to create ties and interact with their peers. Interdependence is present at both levels. At the biological level, interdependence stems from genetic inheritance. At the social level, interdependence emerges from social ties. In this thesis, we investigate whether, in today's highly connected world, individual privacy is in fact achievable, or if it is almost impossible due to the inherent interdependence between people.
First, we study interdependent privacy risks at the social level, focusing on online social networks (OSNs), the digital counterpart of our social lives. We show that, even if an OSN user carefully tunes his privacy settings in order to not be present in any search directory, it is possible for an adversary to find him by using publicly visible attributes of other OSN users. We demonstrate that, in OSNs where privacy settings are not aligned between users and where some users reveal a (even limited) set of attributes, it is almost impossible for a specific user to hide in the crowd. Our navigation attack complements existing work on inference attacks in OSNs by showing how we can efficiently find targeted profiles in OSNs, which is a necessary precondition for any targeted attack. Our attack also demonstrates the threat on OSN-membership privacy.
Second, we investigate upcoming interdependent privacy risks at the biological level. More precisely, due to the recent drop in costs of genome sequencing, an increasing number of people are having their genomes sequenced and share them online and/or with third parties for various purposes. However, familial genetic dependencies induce indirect genomic privacy risks for the relatives of the individuals who share their genomes. We propose a probabilistic framework that relies upon graphical models and Bayesian inference in order to formally quantify genomic privacy risks. Then, we study the interplay between rational family members with potentially conflicting interests regarding the storage security and disclosure of their genomic data. We consider both purely selfish and altruistic behaviors, and we make use of multi-agent influence diagrams to efficiently derive equilibria in the general case where more than two relatives interact with each other. We also propose an obfuscation mechanism in order to reconcile utility with privacy in genomics, in the context where all family members are cooperative and care about each other's privacy.
Third, we study privacy-enhancing systems, such as anonymity networks, where users do not damage other users' privacy but are actually needed in order to protect privacy. In this context, we show how incentives based on virtual currency can be used and their amount optimized in order to foster cooperation between users and eventually improve everyone's privacy. We derive our analytical findings by relying upon Markov chains, game theory, and Markov decision processes. This last part demonstrates that other people can also play a beneficial role in privacy.
We conclude that the quest for online privacy is chimerical because of the lack of individual control over data. As a consequence, unless cooperation between people quickly expands, we should consider that online privacy is steadily vanishing, and start designing novel mechanisms for the upcoming post-privacy era. We should finally redefine privacy, which is, beyond an individual right, now part of the commons.
Mathias Humbert, Erman Ayday, Jean-Pierre Hubaux, Amalio Telenti. On Non-cooperative Genomic Privacy. In Proceedings of the 19th International Conference on Financial Cryptography and Data Security (FC), 2015.
PDF BibTeX Abstract
```
@inproceedings{humbert2015non,
   author      = {Humbert, Mathias and Ayday, Erman and Hubaux, Jean-Pierre
                 and Telenti, Amalio},
   booktitle   = {Financial Cryptography and Data Security},
   title       = {{O}n {N}on-cooperative {G}enomic {P}rivacy},
	year={2015},
	publisher={Springer}
}
                
```
Over the last few years, the vast progress in genome sequencing has highly increased the availability of genomic data. Today, individuals can obtain their digital genomic sequences at reasonable prices from many online service providers. Individuals can store their data on personal devices, reveal it on public online databases, or share it with third parties. Yet, it has been shown that genomic data is very privacy-sensitive and highly correlated between relatives. Therefore, individuals' decisions about how to manage and secure their genomic data are crucial. People of the same family might have very different opinions about (i) how to protect and (ii) whether or not to reveal their genome. We study this tension by using a game-theoretic approach. First, we model the interplay between two purely-selfish family members. We also analyze how the game evolves when relatives behave altruistically. We define closed-form Nash equilibria in different settings. We then extend the game to N players by means of multi-agent influence diagrams that enable us to efficiently compute Nash equilibria. Our results notably demonstrate that altruism does not always lead to a more efficient outcome in genomic-privacy games. They also show that, if the discrepancy between the genome-sharing benefits that players perceive is too high, they will follow opposite sharing strategies, which has a negative impact on the familial utility.
Mathias Humbert, Erman Ayday, Jean-Pierre Hubaux, Amalio Telenti. Reconciling Utility with Privacy in Genomics. In Proceedings of the 13th ACM Workshop on Privacy in Electronic Society (WPES), 2014.
PDF BibTeX Abstract
```
@inproceedings{humbert2014reconciling,
  title={Reconciling {U}tility with {P}rivacy in {G}enomics},
  author={Humbert, Mathias and Ayday, Erman and Hubaux, Jean-Pierre and Telenti, Amalio},
  booktitle={Proceedings of the 13th Workshop on Privacy in the Electronic Society},
  pages={11--20},
  year={2014},
  organization={ACM}
}
                
```
Direct-to-consumer genetic testing makes it possible for everyone to learn their genome sequences. In order to contribute to medical research, a growing number of people publish their genomic data on the Web, sometimes under their real identities. However, this is at odds not only with their own privacy but also with the privacy of their relatives. The genomes of relatives being highly correlated, some family members might be opposed to revealing any of the family's genomic data. In this paper, we study the trade-off between utility and privacy in genomics. We focus on the most relevant kind of variants, namely single nucleotide polymorphisms (SNPs). We take into account the fact that the SNPs of an individual contain information about the SNPs of his family members and that SNPs are correlated with each other. Furthermore, we assume that SNPs can have different utilities in medical research and different levels of sensitivity for individuals. We propose an obfuscation mechanism that enables the genomic data to be publicly available for research, while protecting the genomic privacy of the individuals in a family. Our genomic-privacy preserving mechanism relies upon combinatorial optimization and graphical models to optimize utility and meet privacy requirements. We also present an extension of the optimization algorithm to cope with the non-linear constraints induced by the correlations between SNPs. Our results on real data show that our proposed technique maximizes the utility for genomic research and satisfies family members' privacy constraints.
Mathias Humbert, Erman Ayday, Jean-Pierre Hubaux, Amalio Telenti. Addressing the Concerns of the Lacks Family: Quantification of Kin Genomic Privacy. In Proceedings of the 20th ACM SIGSAC Conference on Computer and Communications Security (CCS), 2013.
PDF BibTeX Abstract
```
@inproceedings{humbert2013addressing,
  title={Addressing the {C}oncerns of the {L}acks {F}amily: {Q}uantification of {K}in {G}enomic {P}rivacy},
  author={Humbert, Mathias and Ayday, Erman and Hubaux, Jean-Pierre and Telenti, Amalio},
  booktitle={Proceedings of the 2013 ACM SIGSAC conference on Computer \& communications security},
  pages={1141--1152},
  year={2013},
  organization={ACM}
}
                
```
The rapid progress in human-genome sequencing is leading to a high availability of genomic data. This data is notoriously very sensitive and stable in time. It is also highly correlated among relatives. A growing number of genomes are becoming accessible online (e.g., because of leakage, or after their posting on genome-sharing websites). What are then the implications for kin genomic privacy? We formalize the problem and detail an efficient reconstruction attack based on graphical models and belief propagation. With this approach, an attacker can infer the genomes of the relatives of an individual whose genome is observed, relying notably on Mendel's Laws and statistical relationships between the nucleotides (on the DNA sequence). Then, to quantify the level of genomic privacy as a result of the proposed inference attack, we discuss possible definitions of genomic privacy metrics. Genomic data reveals Mendelian diseases and the likelihood of developing degenerative diseases such as Alzheimer's. We also introduce the quantification of health privacy, specifically the measure of how well the predisposition to a disease is concealed from an attacker. We evaluate our approach on actual genomic data from a pedigree and show the threat extent by combining data gathered from a genome-sharing website and from an online social network.
Mathias Humbert, Théophile Studer, Matthias Grossglauser, Jean-Pierre Hubaux. Nowhere to Hide: Navigating around Privacy in Online Social Networks. In Proceedings of the 18th European Symposium on Research in Computer Security (ESORICS), 2013.
PDF BibTeX Abstract
```
@incollection{humbert2013nowhere,
  title={Nowhere to Hide: Navigating around Privacy in Online Social Networks},
  author={Humbert, Mathias and Studer, Th{\'e}ophile and Grossglauser, Matthias and Hubaux, Jean-Pierre},
  booktitle={Computer Security--ESORICS 2013},
  pages={682--699},
  year={2013},
  publisher={Springer}
}
                
```
In this paper, we introduce a navigation privacy attack, where an external adversary attempts to find a target user by exploiting publicly visible attributes of intermediate users. If such an attack is successful, it implies that a user cannot hide simply by excluding himself from a central directory or search function. The attack exploits the fact that most attributes (such as place of residence, age, or alma mater) tend to correlate with social proximity, which can be exploited as navigational cues while crawling the network. The problem is exacerbated by privacy policies where a user who keeps his profile private remains nevertheless visible in his friends' ''friend lists''; such a user is still vulnerable to our navigation attack. Experiments with Facebook and Google+ show that the majority of users can be found efficiently using our attack, if a small set of attributes are known about the target as side information. Our results suggest that, in an online social network where many users reveal a (even limited) set of attributes, it is nearly impossible for a specific user to ''hide in the crowd''.
Francisco Santos, Mathias Humbert, Reza Shokri, Jean-Pierre Hubaux Collaborative Location Privacy with Rational Users. In Proceedings of the 2nd Conference on Decision and Game Theory for Security (GameSec), 2011.
PDF BibTeX Abstract
```
@incollection{santos2011collaborative,
  title={Collaborative {L}ocation {P}rivacy with {R}ational {U}sers},
  author={Santos, Francisco and Humbert, Mathias and Shokri, Reza and Hubaux, Jean-Pierre},
  booktitle={Decision and Game Theory for Security},
  pages={163--181},
  year={2011},
  publisher={Springer}
}
                
```
Recent smartphones incorporate embedded GPS devices that enable users to obtain geographic information about their surroundings by providing a location-based service (LBS) with their current coordinates. However, LBS providers collect a significant amount of data from mobile users and could be tempted to misuse it, by compromising a customer's location privacy (her ability to control the information about her past and present location). Many solutions to mitigate this privacy threat focus on changing both the architecture of location-based systems and the business models of LBS providers. MobiCrowd does not introduce changes to the existing business practices of LBS providers, rather it requires mobile devices to communicate wirelessly in a peer-to-peer fashion. To lessen the privacy loss, users seeking geographic information try to obtain this data by querying neighboring nodes, instead of connecting to the LBS. However, such a solution will only function if users are willing to share regional data obtained from the LBS provider. We model this collaborative location-data sharing problem with rational agents following threshold strategies. Initially, we study agent cooperation by using pure game theory and then by combining game theory with an epidemic model that is enhanced to support threshold strategies to address a complex multi-agent scenario. From our game-theoretic analysis, we derive cooperative and non-cooperative Nash equilibria and the optimal threshold that maximizes agents' expected utility.
Mathias Humbert, Mohammad Hossein Manshaei, Jean-Pierre Hubaux One-to-n Scrip Systems for Cooperative Privacy-Enhancing Technologies. In Proceedings of the 49th Annual Allerton Conference on Communication, Control, and Computing (Allerton), 2011.
PDF BibTeX Abstract
```
@inproceedings{humbert2011one,
  title={One-to-n {S}crip {S}ystems for {C}ooperative {P}rivacy-{E}nhancing {T}echnologies},
  author={Humbert, Mathias and Manshaei, H and Hubaux, J-P},
  booktitle={Communication, Control, and Computing (Allerton), 2011 49th Annual Allerton Conference on},
  pages={682--692},
  year={2011},
  organization={IEEE}
}
                
```
Scrip is a generic term for any substitute for real currency; it can be converted into goods or services sold by the issuer. In the classic scrip system model, one agent is helped by another in return for one unit of scrip. In this paper, we present an upgraded model, the one-to-n scrip system, where users need to find n agents to accomplish a single task. We provide a detailed analytical evaluation of this system based on a game-theoretic approach. We establish that a nontrivial Nash equilibrium exists in such systems under certain conditions. We study the effect of n on the equilibrium, on the distribution of scrip in the system and on its performance. Among other results, we show that the system designer should increase the average amount of scrip in the system when n increases in order to optimize its efficiency. We also explain how our new one-to-n scrip system can be applied to foster cooperation in two privacy-enhancing applications.
Mathias Humbert, Mohammad Hossein Manshaei, Julien Freudiger, Jean-Pierre Hubaux Tracking Games in Mobile Networks. In Proceedings of the 1st Conference on Decision and Game Theory for Security (GameSec), 2010.
PDF BibTeX Abstract Data
```
@incollection{humbert2010tracking,
  title={Tracking games in mobile networks},
  author={Humbert, Mathias and Manshaei, Mohammad Hossein and Freudiger, Julien and Hubaux, Jean-Pierre},
  booktitle={Decision and Game Theory for Security},
  pages={38--57},
  year={2010},
  publisher={Springer}
}
                
```
Users of mobile networks can change their identifiers in regions called mix zones in order to defeat the attempt of third parties to track their location. Mix zones must be deployed carefully in the network to reduce the cost they induce on mobile users and to provide high location privacy. Unlike most previous works that assume a global adversary, we consider a local adversary equipped with multiple eavesdropping stations. We study the interaction between the local adversary deploying eavesdropping stations to track mobile users and mobile users deploying mix zones to protect their location privacy. We use a game-theoretic model to predict the strategies of both players. We derive the strategies at equilibrium in complete and incomplete information scenarios and propose an algorithm to compute the equilibrium in a large network. Finally, based on real road-traffic information, we numerically quantify the effect of complete and incomplete information on the strategy selection of mobile users and of the adversary. Our results enable system designers to predict the best response of mobile users with respect to a local adversary strategy, and thus to select the best deployment of countermeasures.

Professional Services

Program committee member of WEIS 2023.
Program committee member of ACM CCS 2023.
Program committee member of ACM CCS 2022.
Program committee member of PETS 2022.
Program committee member of USENIX Security 2021.
Program committee member of CRITIS 2021.
Program committee member of PETS 2021.
Program committee member of PETS 2020.
Program committee member of ACM CCS 2018.
Program committee member of ISMB/ECCB 2018, 2019.
Program committee member of iPAT 2018.
Program committee member of GenoPri 2016, 2017.
Reviewer for IEEE Transactions on Computers, IEEE Transactions on Dependable and Secure Computing, IEEE/ACM Transactions on Networking, IEEE Transactions on Information Forensics and Security, IEEE/ACM Transactions on Computational Biology and Bioinformatics, IEEE Transactions on Knowledge and Data Engineering, IEEE Journal on Selected Areas in Communications, Pervasive and Mobile Computing, Computer Communications, IEEE Security & Privacy Magazine, Knowledge and Information Systems, Computers & Security.

Press

La Télé, Quand nos claviers nous mettent à nu, May 31, 2022.
Le Temps, L’espionnage de nos claviers révélé par un chercheur lausannois, May 16, 2022.
Fortune, Meta, TikTok and thousands of major websites are found swiping data you enter on forms—even if you don’t hit submit, May 12, 2022.
Le Monde, Des milliers de sites Internet enregistrent les données des formulaires, avant même que le bouton « envoyer » ait été utilisé, May 12, 2022.
Wired, Thousands of Popular Websites See What You Type—Before You Hit Submit, May 11, 2022.
Temps Présent (Swiss TV), ADN, réfléchissez avant de cracher, June 27, 2019.
Undark Magazine, The Promise and Perils of Sharing DNA, September 13, 2016.
L'Humanité, Suivi médical, protection des données : quels bénéfices, quels dangers ?, May 3, 2016.

Teaching

Introduction aux Réseaux (Lecturer): Fall 2022.
Technical and Economic Aspects of Cybersecurity (Lecturer): Spring 2022.
Theory of Modern Privacy Research (Lecturer/Advisor): Spring 2016.
Information, Computation, Communication (Guest Lecturer): Spring 2015.
Mobile Networks (Assistant): Spring 2014, 2015.
Pattern Classification and Machine Learning (Assistant): Fall 2013.
Computer Networks (Assistant): Fall 2010, 2011, 2012.
Logic Systems (Assistant): Spring 2011, 2012.

Recent Talks

Machine Learning for Privacy, Privacy for Machine Learning, CUSO doctoral seminar, Neuchâtel, December 2021
Machine Learning for Privacy, Privacy for Machine Learning, University of Bern, November 2021
Machine Learning for Intrusion Detection Systems: Challenges and Opportunities, ZISC Seminar, ETH Zurich, April 2021
Analyzing Cybersecurity Risks with and in Machine Learning, ZISC Seminar, ETH Zurich, April 2021
Enhancing Privacy in Machine Learning, AI and Security Day, CNRS/INSA, Toulouse (France), January 2020
Introduction to the CYD Campus and Privacy, CAS on Data Science and Management, University of Lausanne (Executive Education), November 2019
Introduction to Data Privacy, CAS on Data Science and Management, University of Lausanne (Executive Education), February 2019
Towards Privacy-Preserving Data Science for Personalized Health, CADMOS Day, Biotech Campus, Geneva (Switzerland), November 2018
Data Science and Privacy for Personalized Health, Seminar on Digital Health, IUMSP (CHUV), Lausanne (Switzerland), November 2018
Anonymizing Genetic Data: Are We There Yet?, Future Medicine, Berlin (Germany), November 2018
The Swiss Data Science Center: Its Role and Approach to Data Protection, Seminar on Data Science, CHUV, Lausanne (Switzerland), May 2018
Biomedical Data Privacy: Risks and Protection Mechanisms, TechnoArk Conference on the Quantified Self, Sierre (Switzerland), January 2018
Inferring Social Links from Mobility Profiles, University of Lausanne (Switzerland), November 2017
Reconciling Data Science with Privacy: The Case of Biomedical Data, Swiss Data Science Center, Lausanne (Switzerland), December 2016
New Challenges in Health Data Privacy: The Case of MicroRNA Expression, University of Luxembourg, November 2016
Membership Privacy in MicroRNA-based Studies, CCS, Vienna (Austria), October 2016
Privacy in Epigenetics: Temporal Linkability of MicroRNA Expression Profiles, Summer Research Institute, EPFL (Switzerland), June 2016
Tracking Personal MicroRNA Expression Profiles over Time, EuroS&P, Saarbrücken (Germany), March 2016
Inference and De-anonymization Attacks against Genomic Privacy, ETH Zürich (Switzerland), October 2015
De-anonymizing Genomic Databases using Phenotypic Traits, PETS, Philadelphia (USA), June 2015
Genomic Privacy : De-anonymization, Inference and Strategies, MPI-SWS/Saarland University, Saarbrücken (Germany), June 2015
Genomic Privacy : De-anonymization, Inference and Strategies, CNRS-LAAS, Toulouse (France), June 2015
On Non-cooperative Genomic Privacy, Financial Crypto, San Jose (Porto Rico), January 2015
Interdependent Privacy in the Social and Biological Realms, Cornell Tech, New York City (USA), November 2014