Seeing the Future of Mapping in Crimespotting

sf crimespotting map

This week Stamen Design released San Francisco Crimespotting. It’s a crime map and notification system that allows for time and crime trend analysis. SF Crimespotting has launched just over two years after the release Oakland Crimespotting (Radar post). Stamen had been waiting for crime data all this time and with the launch of DataSF they are able to use an official API for crime data.

SF Crimespotting is very similar to the initial release for Oakland. As I wrote in 2007:

Each type of crime is assigned a color-coded icon with an abbreviation. You can highlight all of a crime type’s markers with a mouseover. You can also change the number of days for which crimes are shown. Each crime has a detail page and that crime can then be viewed in context with others. You can also slice the data by day, type and the intersection of the two. You can also subscribe to get email alerts and RSS feeds for a specific place in Oakland.

The latest releases of the Crimespotting platform reflect several important trends in online mapping:

1) Crowdsourced Maps – When Oakland Crimespotting launched it used Microsoft Live maps (which would now be called Bing). They have switched to Cloudmade maps which are based on Open Street Map data. The maps look amazing and at initial glance they appear to be the same as any other major provider’s maps. Google’s Mapmaker project (Radar posts) has also been seeing more attention and just this week expanded into Mexico (I wonder how long until they bring Mapmaker to the US). Waze (Radar post) is using user-generated traces to create their realtime maps.

pie of time stamen

2) Temporal Mapping – Time is being added to online maps and other visualizations. As data comes to use in realtime there are new conventions that need to be developed. Stamen, through this project and their work with Trulia Hindsight (Radar post) and MySociety (Radar post), are at the forefront of designing methods of dealing with varying scales and types of time data. In their post The Pie of Time Stamen details their thinking for how to represent hours, days and years in the project. The old Crimespotting did not allow you to navigate to archival data. With the new UI there are now permalinks to all crime reports The hours control is shown to the right. Only the crimes that occurred during the highlighted times will appear on the map. Stamen has included quick links to show specific times like “Commute” and “Nightlife”.

slider

The slider and dropdown used to navigate days, months and years are shown above. Each day of the slider shows the total amount of crime that day. The highlighted area dictates the crime shown on the map.

3) Government Data – The new federal administration has shown a renewed interest in releasing data (most of this will have some geocomponent). The 2010 Census is around the corner and that will add to the data flow. As more data is released you can expect an explosion of government mashups. You can also expect more civic minded companies (especially after this week’s exit by Everyblock (Radar post)).

4) Geo-AnalysisGIS used within enterprises, governments and universities are designed to take massive geodata sets and simplify them so that decisions can be made. Crimespotting may look like a slick consumer app and that’s because it is. However as you manipulate its many controls you’ll realize that you can learn a lot about a city and how a time of day or section of the city impacts the likelihood of your being involved in a crime. You can determine if you’re more likely to be mugged in the Castro on Thursdays vs. Tuesdays. The only problem is that you are limited to crime data. I’d love to have ability to add other layers like housing prices or average income. Crimespotting has a read API; I hope Stamen adds Write capabilities.

I’d ask the kind folks at Stamen (very nicely) to make a Crimespotting Seattle, but unfortunately we don’t publicly release our crime data. Here’s to hoping that we get a mayor in this Fall’s election who will open the data coffers. Does your city share its data? If so include a link in the comments.

All of these trends are going to be big topics at this year’s Where 2.0 (3/31-4/2 in San Jose). Submit your topic now!

tags: , , ,
  • http://www.urbanresearch.org Steven Romalewski

    Stamen’s work is very impressive, from cartography to overall design to interactive features.

    But I wanted to comment on your point #4 that the site is limited to crime data and it’d be nice to have other things (house prices, income, etc) to compare to.

    Based on most online mapping applications, I would argue that it’s hard, though not impossible, to do both — Stamen’s approach is quite powerful partly because it’s focused on one “theme” (in this case crime). Expanding this to multiple themes while also providing powerful tools and keeping it simple and easily accessible by a wide audience is, in some ways, the holy grail of going beyond “push pin” mashups.

    I’ve blogged about this a bit at Spatiality, and also discussed it in a paper presented at this year’s GeoWeb conference (here’s the paper in PDF format). We’ve tried our hand at integrating a wealth of mapped data in (hopefully) an easily accessible way at http://www.longislandindexmaps.org and OASIS and are exploring new techniques to integrate the interactive tools that Stamen has provided.

    GeoCommons Maker! has moved far in this arena, though even there the ability to layer in more than two perhaps three data sets can become cumbersome.

    I’m glad that next year’s Where 2.0 includes a topic on “Rich Analysis” — we’re hoping to contribute.

  • http://brookbesor.com Jack Willis

    here’s a link to a short post, with illustration, of a small rural county in Tennessee using a “crime mapping” method/technology to track alcohol and drug abuse among youth, “Hot spot incident mapping in rural tennessee: http://brookbesor.com/blogs/jack/?p=79

  • http://www.metaprinter.com Robert

    I can’t even get my township to set up a police blotter RSS feed

  • personne

    The Tennesee stuff can easily cross the line into a police state. It will also push youth to get more inventive so they can’t be found in their natural explorations.

  • http://spotcrime.com colin

    SF has an alternative crime map at SpotCrime. In addition to the SF PD feed, we add news and scanner data.

  • Ed M

    The US Dept of Justice has some information about Crime Mapping. See “Mapping and Analysis for Public Safety

  • Tols

    Nice. But one could wonder if a criminal would use the map to search for so-believed ‘safe-areas’ for fresh meat.

  • http://www.iyogi.net/ simmondia

    However as you manipulate its many controls you’ll realize that you can learn a lot about a city and how a time of day or section of the city impacts the likelihood of your being involved in a crime.

  • Isak Wøien

    The police should try to keep these ideas some kind of a secret because the whole concept will be a flaw if everybodys awareness about this possibility is out “in the open”. But as a crimeprevention strategy, it should work for at least some time…