Defining a Threshold Value for Maximum Spatial Information Loss of Masked Geo-Data

Abstract

Abstract:Geographical masks are a group of location protection methods for the dissemination and publication of confidential and sensitive information, such as health-and crime-related geo-referenced data. The use of such masks ensures that privacy is protected for the individuals involved in the datasets. Nevertheless, the protection process introduces spatial error to the masked dataset. This study quantifies the spatial error of masked datasets using two approaches. First, a perceptual survey was employed where participants ranked the similarity of a diverse sample of masked and original maps. Second, a spatial statistical analysis was performed that provided quantitative results for the same pairs of maps. Spatial statistical similarity is calculated with three divergence indices that employ different spatial clustering methods. All indices are significantly correlated with the perceptual similarity. Finally, the results of the spatial analysis are used as the explanatory variable to estimate the perceptual similarity. Three prediction models are created that indicate upper boundaries for the spatial statistical results upon which the masked data are perceived differently from the original data. The results of the study aim to help potential "maskers" to quantify and evaluate the error of confidential masked visualizations.

How Is the Confidentiality of Crime Locations Affected by Parameters in Kernel Density Estimation?

“…The confidentiality of crime locations is of concern in many countries. Crime locations could lead to revealing victims' locations, disclosing opportunity information to potential offenders, or even imparting experience to potential offenders [14,38]. To avoid such problems, various types of location protecting techniques have been used to protect these privacies [4,5,9,11,13,14].…”

“…Compared to the city-wide patrol practice, hotspot-targeted patrol strategies could save police resources and deter crime [7,8]. Another benefit of KDE is that it could maintain the confidentiality of crime locations due to its ability to obscure locational information [4,[9][10][11][12][13][14].…”

Confidentiality considerations for use of social-spatial data on the social determinants of health: Sexual and reproductive health case study

“…Selecting an approach to reduce the probability of identifying individuals, while preserving the characteristics of the geographic data for valid inference depends in part on the nature of the data, acceptable confidentiality risk, and current and future use of the data (Armstrong et al, 1999; A. Curtis et al, 2011; El Emam et al, 2011; El Emam et al, 2015b; Kounadi & Leitner, 2015; Seidl et al, 2015; VanWey et al, 2005). For example, point data are more suitable for disease surveillance and outbreak investigation, but have a high risk of compromising individual identity if released publicly.…”

“…Collectively, these findings underscore that despite growing literature raising the alarm about potential confidentiality breaches, as well as development of new methods for geomasking spatial data (e.g., Allshouse et al, 2010; Bader et al, 2016; Gutmann et al, 2008; Hampton et al, 2010; Kounadi et al, 2013; Kounadi & Leitner, 2015; Krumm, 2007; Seidl et al, 2015; VanWey et al, 2005), many public health researchers remain unaware of the potential risks and evolving solutions to help mitigate these risks. Our failure to effectively translate existing knowledge into practice may be due in part to the evolving intersection of two fields of inquiry (e.g., geography and public health).…”

Methods for deriving and calibrating privacy-preserving heat maps from mobile sports tracking application data

“…Because aggregation can hide important spatial patterns in the data, various alternative geo-masking techniques, such as random perturbation and affine transformation (translate, rotate, and scale), have been introduced to preserve the disaggregated, discrete nature of the original data (Armstrong et al, 1999;Kwan et al, 2004). Although they have been used mainly with georeferenced, sensitive health-and crime-related point data (e.g., Leitner and Curtis, 2006;Kounadi and Leitner, 2015), Krumm (2007) and Seidl et al (2015) have applied them also to GPS trajectory data. In this study, where it was crucial to prevent re-identification of an individual user and trajectory while providing the heat map viewer accurate information about popular cycling paths in their actual locations on the road network, geo-masking techniques as such were, however, not adequate.…”