User's involvement in location obfuscation with LBS

Exploring End User Preferences for Location Obfuscation, Location-Based Services, and the Value of Location is an interesting paper written by Bernheim Brush, John Krumm, and James Scott from Microsoft Research. The paper presents the result from a field study about people’s concerns about the collection and sharing of long-term location traces. To do so, they interviewed 32 person from 12 households as part of a 2-month GPS logging study. The researchers also investigated how the same people react to location "obfuscation methods":

  • "Deleting: Delete data near your home(s): Using a non-regular polygon all data within a certain distance of your home and other specific locations you select. This would help prevent someone from discovering where you live.
  • Randomizing: Randomly move each of your GPS points by a limited amount. The conditions below ask about progressively more randomization. This would make it harder for someone else to determine your exact location.
  • Discretizing: Instead of giving your exact location, give only a square that contains your location. Your exact location could not be determined, only that you were somewhere in the square. This would make it difficult for someone to determine your exact location.
  • Subsampling: Delete some of your data so there is gap in time between the points. Anyone who can see your data would only know about your location at certain times.
  • Mixing: Instead of giving your exact location, give an area that includes the locations of other people. This means your location would be confused with some number of other people."

Results indicate that;

"Participants preferred different location obfuscation strategies: Mixing data to provide k-anonymity (15/32), Deleting data near the home (8/32), and Randomizing (7/32). However, their explanations of their choices were consistent with their personal privacy concerns (protecting their home location, obscuring their identity, and not having their precise location/schedule known). When deciding with whom to share with, many participants (20/32) always shared with the same recipient (e.g. public anonymous or academic/corporate) if they shared at all. However, participants showed a lack of awareness of the privacy interrelationships in their location traces, often differing within a household as to whether to share and at what level."

Why do I blog this? Gathering material about location-based services, digital traces and privacy for a potential research project proposal. What is interesting in this study is simply that the findings show that end-user involvement in obfuscation of their own location data can be an interesting avenue. From a research point of view, it would be curious to investigate and design various sorts of interfaces to allow this to happen in original/relevant/curious ways.