2 ms·

>The data, which has been deidentified to protect drivers’ privacy, will allow EFF and ACLU SoCal to learn how the agencies are using automated license plate re

>The data, which has been deidentified to protect drivers’ privacy, will allow EFF and ACLU SoCal to learn how the agencies are using automated license plate reader (ALPR) systems throughout the city and county of Los Angeles and educate the public on the privacy risks posed by this intrusive technology. If the data tells you that much, which it sounds like it does, how hard is it to figure out identities from the patterns?

autoexec 7y ago
With a large enough data set I'm guessing it'd be pretty trivial. It seems like it'd depend on how many places data was collected from. Car #308933938 drove past main street each morning mon-fri between the hours of 7:30am and 7:50am heading north doesn't tell you much, but if you catch the same car at more intersections every morning and don't record it at several other intersections along his projected route you could get a pretty clear picture of where they are going and where they stopped.
shakna 7y ago
De-anonymising data is not especially difficult. [0][1] Combine it with another database or two and you'll have a high degree of confidence you know who did what, where and when. [0] http://digital.law.washington.edu/dspace-law/bitstream/handle/1773.1/417/vol5_no1_art3.pdf http://digital.law.washington.edu/dspace-law/bitstream/handl... [1] https://news.mit.edu/2013/how-hard-it-de-anonymize-cellphone-data https://news.mit.edu/2013/how-hard-it-de-anonymize-cellphone...