Just listened to a recent edition of Radio Open Source on the NSA wiretapping case, and was struck by how well the topic maps to social networks as we know and use them. Data mining, degrees of separation, pattern analysis, and more. With comments by William Gibson!
Apparently it's not about surveillance, it's about mapping social networks. For these large-scale operations the content of each individual phone-call becomes irrelevant; what's more interesting is to find the degrees of separation between everybody, and then to be able to map out interesting subgroups. (See also my Datenspuren 2006 report.)
Patrick Radden Keefe links this new superlative level of data mining with the aftermath of 9/11:
After 9/11, after we learned more about the highjackers involved, some actual network scientists said: We have these 19 highjackers, let's look at who each of them is communicating with, and who shares apartments with whom, and who communicated with whom, and lived with whom, and let's connect them. And what they found was that of those 19 highjackers none of them was separated by more than two degrees of separation, they all had links to another, and a disproportionate number of links converged on Muhammad Atta, as a 'ringleader'. And I think in hindsight there was this idea that "boy if only we'd been able to put those connections together beforehand. If only we'd been able to see those in realtime."
However social network science as used to analyze phone protocols apparently is still in its infancy. And Patrick illustrates another problem in analyzing the phone protocols:
If somebody uses a credit card and buys two dollars worth of gasoline and an hour later buys tens of thousands dollars worth of stuff at an electronics store a red flag goes up for the [credit card] company [...] because that pattern of activity correlates very very closely with instances of fraud. A big problem for the [federal] agencies is that we don't know that much about the patterns of activity of terrorists.
;)
And all the talk about call pattern analysis made me aware that I'm not using my phone lines that much any more, as most of my communication has moved to the 'net. Maybe it's time to move back, for my privacy's sake... (I use Google as a sort of spam filtering email proxy for ~50% of my private email.)
William Gibson in a side-note:
I don't think humanity knows why we're doing any of this stuff. A couple of hundred years down the road, if people look back at what the NSA has done the significance of it won't be about terrorism or Iraq or George Bush or the Constitution, it will be about how we're driven by emerging technologies, and how we struggle to keep up with it.
Comments
More info on social network analysis and looking for patterns in phone calls...
http://www.networkweaving.com/blog/2006/06/un-weaving-networks.html
Valdis Krebs, 2006-07-16 17:19 CET (+0100) Link
Comments are closed. You can contact me instead.