LoPaS '18: Location Privacy and Security Workshop

News

• August 23 Keynote title and abstract are online.
• August 4 Program and accepted papers announced.
• June 4 Deadline for submissions extended to June 15th.
• May 23 Authors of accepted short and full papers will be invited to submit to a Special Feature (Issue) in the Journal of Spatial Information Science (Open call for papers forthcoming).
• May 15 We are excited to annouce that Dr. Benjamin I. P. Rubinstein, Associate Professor, School of Computing & Information Systems, University of Melbourne will give a keynote presentation on topics of privacy and security related to his work in machine learning, deanonymization, and health care medical records.

Scope

Location privacy has been a topic of research for many years but has recently seen a resurgence in interest. This renewed interest is driven by recent advances in location-enabled devices, sensors and context-aware technology, and the broader Internet of Things (IoT). The data generated via these devices are being collected, analyzed, and synthesized at an unprecedented rate. While much of these data are used in the advancement of products or services, many individuals are unaware of the information that is being collected, or how it is being used. The resulting information extracted from these personal data have contributed to significant advances in domain such as location recommendations or fitness/health services, but these advances often come at the cost of location privacy. This workshop is aimed at facilitating a discussion surrounding current methods and techniques related to location privacy as well as the social, political, etc. implications of sharing or preserving location privacy. Further, this workshop invites contributions and discussion related to methods and techniques for securing location information and preserving the privacy of geospatial data.

Topics of Interest

• Context-aware mobile applications
• Obfuscation techniques
• Educational approaches to location privacy
• Policy implications of personal location information
• Role of location in personal relationship development
• Geosocial media implications
• Credibility, trust, and expertise related to location information
• Tools and systems for preserving or securing private information
• Techniques for sharing private location information
• Methods for securing location information
• Place-based data privacy
• Individual vs. group privacy preservation
• Gamification techniques
• Next-generation location-based services
• Geofencing
• Marketplaces for location data

Program

The current schedule for the workshop is as follows. The workshop will take place on RMIT's campus in Building 16 (Storey Hall). The exact room will be posted next to registration in the lobby. Please note that small adjustments may be made as we approach the event.

• 08:45 - 09:00 Welcome
• 09:00 - 10:00 Keynote: Dr. Benjamin I. P. Rubinstein
• 10:00 - 10:30 Lightning Talks / Introductions
• 10:30 - 11:00 Coffee Break
• 11:00 - 12:00 Accepted Paper Presentations
• 12:00 - 12:30 Is Spatial Privacy Special? Introduction of Discussion Topics
• 12:30 - 14:00 Lunch (will be provided)
• 14:00 - 15:30 Break out Groups
• 15:30 - 16:00 Coffee Break
• 16:00 - 17:30 Groups Reporting, Discussion, and Collaborative Planning

Accepted Papers

• Seeking Mr & Ms Regular: Sentinels to Characterize Crowd Dynamics
  Elham Naghizade, Jeffrey Chan, and Martin Tomko
• Exploring the Effectiveness of Geomasking Techniques for Protecting the Geoprivacy of Twitter Users
  Song Gao and Qunying Huang
• trajGANs: Using generative adversarial networks for geo-privacy protection of trajectory data
  Xi Liu, Hanzhou Chen and Clio Andris

Authors of accepted papers have the option of submitting their extended paper to the special issue/feature on Geospatial Privacy and Security in the Journal of Spatial Information Science. The open call for special issue papers is forethcoming. In addition, ALL registered workshop participants are invited to give a 5 minute, ignite style, lightning presentation on a subject related to the workshop topic. To register for the workshop, please visit the GIScience Conference registration page.

Important All workshop participants are expected to adhere to the same code of conduct outlined by the main conference organizers. Please visit the Conference Code of Conduct page for further information.

Important Dates

• Submissions Due: 4 June, 2018 15 June, 2018
• Acceptance Notification: 2 July, 2018 16 July, 2018
• Camera-ready Copies Due: 30 July, 2018
• Workshop: 28 August, 2018

Keynote: Towards Turn-Key Differential Privacy

Speaker: Dr. Benjamin I. P. Rubinstein
Associate Professor, School of Computing & Information Systems, University of Melbourne

Abstract: A decade on, differential privacy has left an indelible mark on the privacy landscape. The framework aims to protect the privacy of participants in a dataset, individually, while releasing statistics to untrusted third parties in aggregate. Actively researched across CS theory, security, databases and learning, differential privacy has shipped in deployments at Google, Apple and Uber. In 2017, Dwork et al. were awarded the Gödel Prize for introducing the framework. Despite these successes, a pain point for practitioners is the need to derive certain sensitivity bounds. After covering basic definitions I will review some common generic mechanisms for privatization which (randomly) smooth non-private releases based on the level of Lipschitz smoothness of the non-private release function: more sensitive analyses requires more smoothing to achieve the same level of privacy. I will present a new sensitivity sampler (Rubinstein & Aldà, ICML’17), which applies empirical process theory to estimate sensitivity generically, even of black-box software programs, obviating manual mathematical analysis (and delivering an increase to utility), at the cost of a slight weakening of privacy guarantees. I will demonstrate these algorithms via open-source R package diffpriv.

Biography: Dr. Benjamin Rubinstein actively researches topics in machine learning, security & privacy, databases such as adversarial learning, differential privacy and record linkage. Prior to joining the University of Melbourne in 2013, he enjoyed four years in the research divisions of Microsoft, Google, Intel and Yahoo! (all in the United States), followed by a short stint at IBM Research Australia. As a full-time Researcher at Microsoft Research, Silicon Valley, Dr. Rubinstein shipped production systems for entity resolution in Bing and the Xbox360; his research has helped identify and plug side-channel attacks against the popular Firefox browser, and deanonymise an unprecedented Australian Medicare data release, prompting introduction of the Re-identification Offence Bill 2016. Since joining Melbourne in 2013, he has led $2.0m in awarded competitive funding ($1.2m per CI). His work has been recognised through an Australian Research Council DECRA award, and a Young Tall Poppy Science award.

Organization Team

• Grant McKenzie, McGill University, Québec, Canada
• Carsten Keßler, Aalborg University, Copenhagen, Denmark
• Clio Andris, Penn State University, State College, USA

Program Committee

• Marc P. Armstrong, University of Iowa
• Carson Farmer, University of Colorado, Boulder
• Sebastien Gambs, Université du Québec à Montréal
• Yingjie Hu, University of Tennessee, Knoxville
• Krzysztof Janowicz, University of California, Santa Barbara
• Peter Johnson, University of Waterloo
• Bernd Resch, University of Salzburg
• Colin Robertson, Wilfrid Laurier University
• Dara Seidl, San Diego State University
• Martin Tomko, University of Melbourne