2008.01.25
First Roanoke City crime map update
The Roanoke City crime map has its first update, with data through January 19. At the same time, to keep the map from getting too cluttered, I removed the oldest offenses. So, there's now exactly a month's worth of reports mapped.
Also, if you want to know what each offense type listed means, you can check out the definitions from the Federal Bureau of Investigations.
I wish I had a convenient way to archive these maps, but I don't -- not yet. But we're still plugging towards building a crime mapping site that has a searchable archive. One day, hopefully soon, you'll be able to sort the map by offense, limit it by a date range, or by offenses within a certain distance of an address you enter.
For now, we're still on the low-tech end of this high-tech kind of undertaking.
You can read more of my thoughts on crime maps in Sunday's paper, by the way. Look for my next DataSphere column in the Virginia section.
"offenses within a certain distance of an address you enter."
That would put this map far ahead of anything else you could find. This is really what people want to know. How close is criminal activity to me?
I'd really, really like to see you archive and then be able to find what and where crime is increasing or decreasing. I bet the police statistician can get this information. How else can they adequately manage their resources?
Comment by Valerie — January 25, 2008 @ 9:34 pm
I've posted the Fairfax County PD website before that does incident mapping, and allows searching by address and radius:
http://www.fairfaxcounty.gov/gisapps/myneighborhood/mynpolice.aspx?fxmResolution=1024x768
However, I read the "Terms and Conditions" more closely this time, and found that it isn't as accurate as I initially thought. Many crimes are not listed, understandably to protect the victim (they list rape and sexual assault, though some are listed in a grouped manner). In addition, it also lists only the initial report received by the 911 center, or as determined by the responding officer. In effect, someone could have called about a burglary (which would be entered on the map), but the responding officer found it to be a property destruction caused by a kid breaking a fence to retrieve his ball (may not be entered).
I haven't found a comprehensive listing of crimes on a map in Fairfax yet, but they do publish weekly activity by police district so it seems there should be a way to make an accurate map.
I would think it's a matter of making a database with these weekly reports, being able to run a query and populating the results on a Google map. Of course, "easy" meaning I'm not the one who has to do it.
Another note (more a word of caution): During our neighborhood watch training, the CRO noted to us that they can only map the crimes that have been reported. Some areas of higher crime don't get as many reports because of whatever reason (residents afraid of retaliation, etc.). Some areas of lower crime may be reported more (such as a neighborhood watch just starting up).
That partially relates to your comment, Valerie. The CRO did note that if they aren't receiving reports, then they must assume that all is fine and focus resources in other areas. I'm sure Roanoke has a similar way of doing things.
Comment by Ed S. — January 26, 2008 @ 10:17 am
Right, Ed. Not only is a map only as good as the data behind it, as I said in my column in the paper today, but the data is only as good as what gets reported.
So, as you point out, a neighborhood with an active community group that calls the police more frequently could appear to have more crime than others.
Comment by Matt — January 27, 2008 @ 10:45 am