Decentralized geoprivacy: leveraging social trust on the distributed web

Abstract

Despite several high-profile data breaches and business models that commercialize user data, participation in social media networks continues to require users to trust corporations to safeguard their personal data. Since these data increasingly contain geographic references that allude to individuals' locations and movements, the need for new approaches to geoprivacy and data sovereignty has grown. We develop a geoprivacy framework that couples two emerging technologies -decentralized data storage and discrete global grid systems -to facilitate fine-grained user control over the ownership of, access to and map-based representation of their data. The framework is illustrated with a dynamic k-anonymity model that links the geographic precision of shared data to social trust within in a social network. In this framework, users' spatio-temporal data are shared through a decentralized system and are represented on a discrete global grid data model at spatial resolutions that correspond to varying degrees of trust between individuals who are exchanging information. Our framework has several advantages over centralized geoprivacy approaches, namely trust in a third-party entity is not required and geoprivacy is dynamic and context-dependent with users maintaining autonomy. As the distributed web begins to emerge, so too can the next generation of geographic information sharing tools.

PrivyTo: A privacy‐preserving location‐sharing platform

“…For instance, Polzin and Kounadi (2021) propose an adaptive technique for residential masking, while Rao et al (2020) developed an innovative privacy‐focused approach to protecting trajectories. Other techniques such as verified neighbor (Richter, 2018) or social trust (Hojati, Farmer, Feick, &, Robertson, 2021) use a modified k ‐anonymity approach. k ‐anonymity is a technique that suggests that by combining a similar set of observations, identifying information about a specific individual, or location, can be obscured, while the data still remain useful.…”

MapSafe: A complete tool for achieving geospatial data sovereignty

“…Obtaining sovereignty over geospatial data, however, is a multifaceted challenge. Current geoprivacy approaches have only addressed one or more of these challenges: focusing on masking (Ajayakumar et al, 2019; Hojati et al, 2021; Swanlund et al, 2020), hexagonal binning (Hojati et al, 2021; Rao et al, 2021), encryption (Hojati et al, 2021), or notarisation on the blockchain (Farnaghi & Mansourian, 2020; Sladić et al, 2021; Zhang et al, 2020), and/or those that offer multi‐level geomasking with encryption (Hojati et al, 2021). MapSafe leverages all these security functions in a workflow to offer a more holistic and complete geospatial data sovereignty approach.…”

“…Data sovereignty solutions are incomplete without giving SDOs the ability to exercise their rights and control the usage and sharing of their data (Hojati et al, 2021; Zhang et al, 2020). As such, accommodating sharing of potentially sensitive data with different sets of users based on varying degrees of trust is a key prerequisite and encourages collaboration between otherwise isolated members.…”

“…As such, accommodating sharing of potentially sensitive data with different sets of users based on varying degrees of trust is a key prerequisite and encourages collaboration between otherwise isolated members. Geospatial researchers have recently emphasized that “fine‐grained control over the resolution of spatio‐temporal data being shared and with whom they are sharing is not provided by most geoprivacy approaches, which tend to rely on simple binary classifications” (Hojati et al, 2021, p. 2). The authors propose a geoprivacy solution that gives social media users control over how their locations are disclosed (higher or lower precision) to their connections based on social trust, namely, the length of their friendship on Twitter.…”

Towards geospatial blockchain: A review of research on blockchain technology applied to geospatial data

“…Location obfuscation mechanisms are commonly used in the models, which aim to protect privacy by deliberately degrading the precision of location information in a way to hide the sensitive information with the precondition that the service can still be conducted to some acceptable extent (e.g. Partovi et al, 2020;Zurbarán et al, 2020;Hojati et al, 2021). For example, Zurbarán et al (2018) developed an algorithm called Rand-K to minimize the impact of location obfuscation on exploratory spatial data analysis (ESDA).…”