Geocoding and reverse geocoding have raised potential privacy concerns, especially regarding the ability to reverse engineer street addresses from published static maps. By digitizing published maps it is possible to georeference them by overlaying with other spatial layers and then extract point locations which can be used to identify individuals or reverse geocoded to obtain a street address of the individual. This has potential implications to determine locations for patients or study participants from maps published in medical literature as well as potentially sensitive information published in other journalistic sources.
In one study a map of Hurricane Katrina mortality locations published in a Baton Rouge, Louisiana, paper was examined. Using GPS locations obtained from houses where fatalities occurred, the authors were able to determine the relative error between the true house locations and the location determined by georeferencing the published map. The authors found that approximately 45% of the points extracted from the georeferenced map were within 10 meters of a household's GPS obtained point.4 Another study found similar results in examining hypothetical low and high-resolution patient address maps similar to what might be found published in medical journals. They found approximately 26% of points obtained from a low-resolution map and 79% from a high-resolution map were matched precisely with the true location.5
The findings from these studies raise concerns regarding the potential use of georeferencing and reverse geocoding of published maps to elucidate sensitive or private information on mapped individuals. Guidelines for the display and publication of potentially sensitive information are inconsistently applied and no uniform procedure has been identified. The use of blurring algorithms which shift the location of mapped points have been proposed[by whom?] as a solution. In addition, where direct reference to the geography of the area mapped is not required, it may be possible to use abstract space on which to display spatial patterns.
Danalet, Antonin; Farooq, Bilal; Bierlaire, Michel (2014). "A Bayesian approach to detect pedestrian destination-sequences from WiFi signatures". Transportation Research Part C: Emerging Technologies. 44: 146–170. doi:10.1016/j.trc.2014.03.015. http://infoscience.epfl.ch/record/199471 ↩
Google Codesource reverse Geocoding API https://code.google.com/apis/maps/documentation/services.html#ReverseGeocoding ↩
OSM Search Engine https://wiki.openstreetmap.org/wiki/Search_engines ↩
Curtis, A. J., Mills, J. W., & Leitner, M. (2006) Spatial confidentiality and GIS: re-engineering mortality locations from published maps about Hurricane Katrina J Health Geogr, 5, 44. http://www.ij-healthgeographics.com/content/5/1/44 ↩
Brownstein, J. S., Cassa, C. A., Kohane, I. S., & Mandl, K. D. (2006) An unsupervised classification method for inferring original case locations from low-resolution disease maps Int J Health Geogr, 5, 56 http://www.ij-healthgeographics.com/content/5/1/56 ↩