Can You Hear Me Now…
Probably the best wrap up on today’s revelations of spying on US citizens is at the Maha blog:
Later in the day there will be more reaction from the Kool-Aiders, but so far we’ve got the NSA story must be a lie. This is coupled with “I bet a little research into the political leanings of the people who are making these claims will ferret out a motive.” Protecting the Fourth Amendment has become subversive.
Stephen Spruiell at NRO assures us that “this is not an eavesdropping program,” so what are we worried about? And cleverly anticipating objections from us liberals, Spruiell adds, “Data mining programs like this one might or might not be effective tools in the war on jihadists, but one thing we know for sure is that the left will not be joining us in a rational debate.” Be advised that any objection we lefties raise will be, by definition, irrational. On the other hand, it is perfectly rational to accept whatever the government tells us without question. They wouldn’t lie to us, right? Those righties are always so logical.
But a first a word from internationally renowned mathematical scholar Hindrocket:
Two, it’s obvious that what the NSA does with this vast amount of data is to run it through computers, looking for suspicious patterns, especially involving known or suspected terrorist phone numbers. I did a quick calculation: assuming that there are 200 million adult Americans, each of whom places or receives ten phone calls a day (a conservative estimate, I think), it would require a small army of 35,000 full-time NSA employees to pay a total of one second of attention to each call. In other words, lighten up: the NSA obviously isn’t tracking your phone calls with your friends and relatives.
Oi.
No, it’s not at all obvious that NSA uses “known or suspected terrorist phone numbers.” (Yes, operator, do you have a listing for Ima Terrorist, Jihadish Bent, Indiana? ).
What’s obvious to me is that they are using a pattern recognition technique called Bayesian Nets. In this case, I would venture to guess that each node on the network represents a phone number. What’s called an edge of the network, the line that connects the nodes, represents which node talks to which other node. The observed inter-relationship of nodes is compared to some known model. In this case we could compare it to a calibrated graph of nodes and edges observed in, for example, organized crime. (Why aren’t we comparing it to a graph of terrorists? I’m assuming that we don’t have enough data to have confidence in that result.) We can also compare it to a graph taken from members of a Girl Scout Troop. Then we ask: is it more likely that the observed graph came from a terrorist sleeper cell (which we model by the organized crime measurements) or from Girls Scouts?
The point here is that this requires creating graphs using everyone’s calls. Even yours, Hindrocket. The result would be a database of graphs showing who associates with who.
Getting the picture?
May 11th, 2006 at 7:18 pm
What’s obvious to me is that they are using a pattern recognition technique called Bayesian Nets. In this case, I would venture to guess that each node on the network represents a phone number. What’s called an edge of the network, the line that connects the nodes, represents which node talks to which other node. The observed inter-relationship of nodes is compared to some known model. In this case we could compare it to a calibrated graph of nodes and edges observed in, for example, organized crime. (Why aren’t we comparing it to a graph of terrorists? I’m assuming that we don’t have enough data to have confidence in that result.) We can also compare it to a graph taken from members of a Girl Scout Troop. Then we ask: is it more likely that the observed graph came from a terrorist sleeper cell (which we model by the organized crime measurements) or from Girls Scouts?
Did I just fall into an episode of Numbers? That would be okay with me. I love that show. Mmmmmm … math …
But beating up on Hindrocket and the rest of them using these kind of rational, informed arguments is just unfair.
May 11th, 2006 at 10:41 pm
Poor Hindrocket. Maybe I’ll send him a gift to show there’s no hard feelings. A bucket of wingnuts, perhaps?
May 12th, 2006 at 3:47 pm
♫ Get a bucket of wingnuts (finger-lickin’ good!), have a barrel of fun! ♫