A Radical Approach to Open Police Data

PATF_Final_Report_Executive_Summary_4_13_16On page 112 of the report of the Chicago Police Accountability Task Force, “Recommendations for Reform: Restoring Trust between the Chicago Police and the Communities they Serve“, there is a section about the publication of police data that might change the way we think about “crime data” in the #opengov and #civictech movement.

In the section on Early Intervention & Personnel Concerns, the main recommendation for this area is that the Chicago Police Department embark on the “design and implementation of a mandatory EIS that centrally collects data across a broad range of data points to capture information on the totality of officer activity”. This kind of system is typically called an “Early Intervention System”, or EIS.

But later in this section is a recommendation that is near and dear to my heart (disclosure: I served on the Early Intervention & Personnel Concerns Working Group for the Task Force).

Here’s the nub:

Continue reading “A Radical Approach to Open Police Data”

Evidence of Ascension in America, Even When Only Half of Us Are Looking

I listen to Fox News on my Sirius XM radio while driving to and fro from Winfield. I consider it something of an intelligence-gathering  mission from the fact-based world, the one where we seek out the truth and take steps to understand it. I also seek to understand the roughly half of my country that aligns itself behind the thin theories that ring there.

So I know the special place that the Benghazi attacks have for the people at Fox News. For them, Benghazi is everything. It is the alpha and the omega of the Obama administration. It is the most popular player in their Fantasy Impeachment league, the one that exists in their stunted minds.

So I was stunned and surprised to see so much ink, and what appears to be such an enormous amount of reporting time, that went into the NYT reporting on A Deadly Mix in Benghazi. It is some of the most thorough, detailed, cited, minute-by-minute tick-tocks I’ve ever read in those pages. It’s unimpeachable, so to speak.

And no one in NYT-Land cares at all about this story. It is not in the top 20 viewed on emailed stories on the site. Nobody cared out there.

It’s a top story on Fox News, though. Because if this story is right, half of their news is wrong. Half of their editorial policy is wrong. So they spend a lot of words doing shoddy followup (unidentified witnesses, compound sentences with indeterminate noun clauses, etc.)

The cool thing is that nothing in the Fox report contradicts the reporting of the NYT whatsoever. They say it was a coordinated attack, and so does the NYT (their timeline starts with photographic surveillance hours before the disturbance). They say “The bosses on the ground were pointing, commanding and coordinating”, and so does the NYT.

Whenever they disagree on a point (like whether the main suspect was affiliated with Al Qaeda), the NYT has dozens of documented interviews with the suspect and people who have known him since he was a child. People who have lived with him in prison.  And Fox has an unidentified source saying “…There is literal evidence in many forms and shapes, directly linking him.”

Literal evidence!

I feel the momentum of America. It’s evident everywhere. In gay rights legislation, in Boehner when asks about who’s kidding who, even in the acquiescence to a Pope who demands we love one another. Everyone.

Evidence everywhere.

A Look at the Databases Used in Reporting “City gives felon six-figure grant to open liquor store”

This morning I read the story, “City gives felon six-figure grant to open liquor store: Liquor store with ties to felon just the latest problem for a blighted neighborhood that has suffered for decades” by Chicago Tribune reporters David Jackson and Gary Marx. I was struck by two things: the complexity of the issue of economic development in an area that has been under-capitalized for decades, and the large number of databases that the reporters used to tell the story.

I have little to say about the first topic, except to again note its complexity. Presumably it is a good thing to stimulate economic investments by people in their own neighborhoods. Presumably liquor is a popular consumer good for large swaths of Chicagoans (as a recovering alcoholic, I can attest to that). Presumably there are many convicted felons who live in neighborhoods all over Chicago. And presumably there are many gangs all over this city who engage in illegal activities. The issues of addiction, incarceration, and gangs are not isolated to the neighborhood referenced in the story, and I think that drilling down into a specific example of these layered issues, without an examination of the root causes, can be problematic.

Having said that, the story is scrupulously reported and fascinating in its detail, and that’s what I want to cover here. In reading the story, I can just see the databases behind it. The story is not presented as “data journalism”, which has been a hot and growing specialty and a field in which I took part for some time, but it is data, and it is journalism.

Continue reading “A Look at the Databases Used in Reporting “City gives felon six-figure grant to open liquor store””

Toward a Generic Context Engine for Civic Data

I have been thinking and writing for quite some time about how to use civic data— the type of info found on data portals like data.cityofchicago.org— to tell real stories. That’s why I’m psyched to be attending two events— Knight-Mozilla-MIT “Story & Algorithm” Hack Day on Saturday and the MIT – Knight Civic Media Conference, where I’ll be on a panel called, “Turning Data Into Narrative“.

Currently, turning data into narrative is a highly hand-sewn effort. Take, for example, this post about a neighborhood building being turned into a Walgreens I compiled this from ten databases, including building permits, business licenses, historic preservation lists, Sanborn maps from 1929 and 1950, county assessor, and county recorder of deeds. None of these databases are in one place, and none of them know how to consume or interpret information from the other. I also pinched info from five news sources, including an article from the New York Times archive about a bank run at this location in 1931.

What if there was a machine where I could enter a piece of information (the name of a lobbyist, the address of a building, the name of a construction project) and pick from a list of data sources (like those listed above) and teach the machine what the data means? Why can’t I get that machine to bring me back a list of businesses that used to be in that building (from municipal business license records and private landlord rent rolls) or the number of babies born there (from birth certificates) or what those babies grew up to do in this world (from the obituaries of the future)? OK, that last item might be hard. But I want this machine.

I’ve got some ideas on specific manifestations of this machine, and I hope to write about them this week as I prepare for Boston.

Here’s some posts wherein I drill down into data and tell a cohesive narrative about the place:

The grand-daddy of all of my efforts in this regard is a screed called, “New Walgreens in the Old Noel State Bank Building at 1601 N. Milwaukee (corner of North, Milwaukee, and Damen in Bucktown/ Wicker Park)“. It’s still not completed– a narrative in progress!