Menu

UCL Department of Geography

Home

Description Photo Here

Personal tools
Log in
This is SunRain Plone Theme
UCL Home  /  Geography  /  News & Events  /  News  /  News Archive  /  August 2015  /  Privacy-preserving techniques for GPS data mining

Privacy-preserving techniques for GPS data mining

New UCL Geography Research

Privacy-preserving techniques for GPS data mining

In a paper this month in EPJ Data Science (Springer Open Access Journal), Luca Rossi (now at Aston), James Walker (now at Mitsubishi UFJ Securities) and Mirco Musolesi (UCL Geography) investigate the risks associated with user identification by Global Positoning System (GPS) datasets, and discuss the design of privacy-preserving techniques for GPS Data Mining.

One of the greatest concerns related to the popularity of GPS-enabled devices and applications is the increasing availability of the personal location information collected through them and shared with application and service providers. More specifically, companies such as telecommunication operators and service providers and government organizations have access to large collections of person and communication data. These may be used, for example, to maintain and manage communications, marketing, security and surveillance services. They include person location data collected from GPS devices, cellular phone usage and WiFi hotspots.

The paper presents a series of techniques for identifying individuals from their GPS movements. More specifically, it includes a detailed analysis of the discriminatory power of speed, direction and distance of travel. A simple yet effective technique is also presented for the identification of users from location information not included in the original training dataset, thus raising important privacy concerns for the management of location datasets. Finally a method is proposed to measure the extent to which a dataset can resist an identification attack based on the techniques described in the paper.

These results have important practical implications, given the increasing interest in mining location information in academia, industry and government. In particular, Mirco Musolesi’s group in UCL Geography is interested in analysing location datasets for both the study of human dynamics at different geographic scales, and the development of practical applications, for example employing behavioural data analytics and designing intelligent location-aware services and systems.

See:

 

EPJ Data Science paper (Open Access): http://www.epjdatascience.com/content/pdf/s13688-015-0049-x.pdf


Image

Trajectory-based identification


Navigation
Tweets from @UCLgeography
@UCLgeography