Got the data, now what?

[Casey Cora]

Our CAR ace has put together a preliminary framework in Django, I've been collecting data on homicides in our area since the first of the year. We have several categories (age, race, address, homicide location, etc. ) and subcategories (weapon, weapon type, relationship to suspect/victim, etc.) and so far, so good.

The idea is to build a page about homicides, and I've got tons of good story ideas because of it, both for print and multimedia.

But I'd to hate make all that data - all that rich, rich data - go to waste in some simple mashup map. Problem is, beyond number-crunching trend stories, I'm kind of at a loss on what that's going to be.

Anyone have any ideas?

[Brad King]

Well, first thing is you absolutely create a map - and you also make sure the information is tagged by zip code and type of crime and whether it was solved or not. Then, you need to get information like property values in areas, money spent on cops for different areas, legislation that was passed by the city and where the money was spent, population decline, where campaign contributions came from, educational test scores...

That type of map would allow you to see some amazing data sets that would probably illuminate problems that we know about.

Then you open up the API for people -- and you make sure the data is sortable for your readers. Give them the tools to make their own mashups and add their own data fields.

And then I would set about getting people to submit stories or pictures (flickr, tagged) or video (Google/YouTube, tagged) with a zip code in the tag as well so you could create real stories from people. You'll need to find some incentive for people to contribute -- and likely you'd need someone to manage that information as it comes in -- but the idea is to spotlight a problem in communities and let people tell their stories.

If you city and county commissioners run in districts, I would have a list of who they are mapped on the map of the other data areas so people could cross reference what their elected officials are doing about the problem.

And I'd have a MeetUp function to do a town hall meeting about it -- see what, if anything, comes out of it -- and use that to get audio and video from folks telling their stories.

Those are just a few things that come to mind. (And I'd get rid of computer assisted reporters and make everyone collect data :)

[Casey Cora]

Thanks a lot. That's a helluva start and some mighty fine ideas. I'm afraid it's gonna take more manpower than just me, though. (not that I'm not up for it.)

It's my first stab at this type of thing, so I wanna knock it out of the park.

[Brad King]

Haha. Not the first time I've been accused of over-reaching. At the very least, you should go look for existing databases of information -- The US Census, Thomas (Library of Congress legislation) -- anything you can readily access already.

The thing is: newspapers can't gather all of this information because they are too diversified. But they can go access the data that organizations dedicated to tracking information already have.

For instance:

Follow the Money: The Nation's Most Complete Resource for Information on Money in State Politics

or

Open Secrets

In other words, don't be intimidated by not having data. Seek out the data that people already have. Once you do that -- and you create wicked awesome stories and interactive maps with it -- the case for collecting more localized information should be easier.

[Brad King]

Here's an example of another group that's using crime data to create a visual display of stories: SpotCrime.

Tie that in with something that I saw at SXSW, Outside.In, which geo-located blogs based upon zip code.

You can begin to see a way to build a community-engaged network that could be used to build a very cool neighborhood information tool.