Tuesday, June 11, 2013

Big Data in Chicago

This is kind of interesting in its own right, and maybe clarifying in some other domains, on the precipitous decline in murder rates in Chicago over the past year, and what the cops hope is causing it:
In recent months, as many as 400 officers a day, working overtime, have been dispatched to just 20 small zones deemed the city’s most dangerous. The police say they are tamping down retaliatory shootings between gang factions by using a comprehensive analysis of the city’s tens of thousands of suspected gang members, the turf they claim and their rivalries. The police also are focusing on more than 400 people they have identified as having associations that make them the most likely to be involved in a murder, as a victim or an offender.
This is describing a Big Data approach to policing—the use of social network analysis to decide where to focus police attention—and it strikes me that it's in a way the opposite of the sort of Big Brother policing we are all worrying about in connection with the National Security Agency and total information awareness and so forth. [jump]
Copies of Indira Freitas Johnson's Peace piece in Chicago for the citywide Ten Million Ripples project. Loyola Dunes Restoration.

Suppose we take the Stop 'n' Frisk system of Vinegar Ray Kelly (as opposed to his benign previous self, the Sugar Ray Kelly of 20-odd years ago) as representing the real Panopticon, in which some number of arbitrary persons within your community is being watched at all times, so that you can never be certain you're not being watched yourself—with the intention that you should therefore not have a gun in your pocket (sounds good!), or smoke weed (sounds—uh, legal), or make a lot of embarrassingly loud noise, or say anything disrespectful about the authorities (oh, wait a minute). What it does is to start from data points, the parameters of profiling (I know racial profiling is no longer legal, but [1] there are ways around that, and [2] don't make me laugh), and target individual people, your potential criminals, who must evade you or avoid breaking the law.

The social network system, in contrast, starts from people and targets the data points: places, metaphorically, in the network, and literal places in the neighborhood, where the means of violence are most densely located—with the intention that the most predictable crimes should not take place, and that the person there should be neither a criminal nor a victim. The cops are not there to cow anybody, not even from smoking weed (they'll be more active, no doubt, where weed is sold—more guns); they're there to inhibit harm from happening. And maybe—if only in this one case, of Chicago gang homicide—it works. (As did the small-data equivalent, the old Dinkins-Kelly program of community policing, which is where New York's murder rate began its decisive fall.)

Rooting out subversion from Napoleon to Onkel Ulbricht has always been about targeting people, profile-provoke-sting; infiltrating their clubs and getting to know their wicked secret thoughts; hemming them round with a secret network of your own. And we don't want that. I especially don't want New York City cops lurking in disguise around mosques and halal butcher shops and sneaking into Muslim students' camping trips. Not that they shouldn't get to know the neighborhood, but they need to get to know it in the first place as something other than a war zone. And then of course they need to let subversion happen and stop actual harm.

Social network analysis is subject to abuse too, though I wouldn't count as abusive the example I linked to above, where the British use computers to stop Paul Revere (a revolution really isn't a dinner party, and you shouldn't start one if you object to being shot at). It's abusive when the links followed are those of Code Pink and the New Black Panthers and the Society of Friends. But the way they stopped the plans of Najibullah Zazi in 2009, the only plot of recent years that definitely wasn't being run by FBI informants, was a pretty good approach, even if our congresspersons can't stop lying about it.
Tucking the Buddha heads into pockets of city neighborhoods that would not ordinarily inspire transformative thought among citizens is part of the project’s goal to beautify the city, Ms. Johnson says. But it’s also to jolt people into changing the way they think about their environment.... it serves a different purpose “when you have it in a marginalized space that isn’t so lovely.”

No comments:

Post a Comment