Open Place Database

54 views
Skip to first unread message

Justin York

unread,
Dec 11, 2013, 6:23:12 PM12/11/13
to root...@googlegroups.com
While developing Find-A-Record, we realized we needed the ability to geocode our data to shapes (place boundaries) as opposed to just coordinates. But that wasn't possible with any of the available services (OSM, Google, Geonames, etc) so we built something which can do it. We named it the Open Place Database (OPD).

Goals for OPD:

* Open data with the Open Database License (OBdl)
* Contain data for the whole world
* Contain historical data as opposed to just current

OPD is not designed to enable standardizing; it's designed to enable geocoding (and perhaps reverse geocoding in the future).

Questions? Concerns? Do you know of any juicy data sets we can get our hands on?

Ben Brumfield

unread,
Dec 11, 2013, 6:29:10 PM12/11/13
to root...@googlegroups.com
This looks awesome!  I'd pop over to the (also relatively new) GeoHumanities SIG and ask on their mailing list: http://geohumanities.org/?q=about_intro

Ben


--
 
---
You received this message because you are subscribed to the Google Groups "rootsdev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rootsdev+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

todd.d....@gmail.com

unread,
Dec 11, 2013, 9:03:53 PM12/11/13
to root...@googlegroups.com
An exciting development Justin! What does your schema look like? I'm especially curious how you are encoding your date ranges.

Also, have you followed the Open Historical Map at all?[1] I think there is a lot of interesting work to be done in this area of open mapping.

Cheers!

--
 
---
You received this message because you are subscribed to the Google Groups "rootsdev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rootsdev+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


--
Tod Robbins
Digital Asset Manager, MLIS

todd.d....@gmail.com

unread,
Dec 11, 2013, 9:07:06 PM12/11/13
to root...@googlegroups.com
PS: Mapbox's iD editor, that they built for OSM, is open source and would be a great fit for an editor on OPD, especially it's ease of use. Is your object model GeoJSON?

todd.d....@gmail.com

unread,
Dec 11, 2013, 10:05:26 PM12/11/13
to root...@googlegroups.com
Ah, I could've just read the About page all the way through. Ha! What are your data sources for this?

Ben Laurie

unread,
Dec 12, 2013, 3:03:13 AM12/12/13
to root...@googlegroups.com
Nice! I already have a request: make it possible to view more than one
era at once...

Openstreetmap has modern boundaries, of course:
http://wiki.openstreetmap.org/wiki/Key:boundary.

Dovy Paukstys

unread,
Dec 12, 2013, 12:42:43 PM12/12/13
to root...@googlegroups.com
WTG Justin. I'll see if maybe we can contribute in some way. Great work!

Robert Hoare

unread,
Dec 12, 2013, 1:43:54 PM12/12/13
to root...@googlegroups.com
That was quick Justin!  I'm surprised you didn't get more responses to your question over at SE GIS: http://gis.stackexchange.com/questions/79510/how-can-i-geocode-to-a-shape-instead-of-a-coordinate/79519 - did you do the demo app with the US counties GeoJSON set I mentioned, or some other source?  It would certainly help to show your sources, especially for historical data where they may vary and where it's hard to get data with a commercial-use license.

You're encouraging me to get on with releasing a minimum viable product at my openplacenames.com where I'm working on something similar from a different angle.  I tend to wait until I'm happy that something is usable before releasing and announcing, but maybe just getting a small subset of it out there and getting feedback is a better approach.

Rob


On Wednesday, December 11, 2013 4:23:12 PM UTC-7, Justin York wrote:

Justin York

unread,
Dec 12, 2013, 4:01:42 PM12/12/13
to root...@googlegroups.com
We loaded a subset of the data from the Atlas of Historical County Boundaries. The license is more restrictive than we would like, but we're okay to use it for a proof of concept as we have now. If we don't get approval then there are some other places to look. The worst that happens is we have to re-create it ourselves by tracing maps which we'll have to do eventually anyways for most parts of the world so we're prepared to tackle that problem.

Robert, we didn't use the data source you mentioned because we wanted historical boundaries too.

What are your plans with openplacenames? How will it be different than our project?


--

Brooke Ganz

unread,
Dec 12, 2013, 10:14:53 PM12/12/13
to root...@googlegroups.com
This is great!  I unfortunately don't have any data sources to suggest at the moment, so please accept some rah-rah enthusiasm instead.  :-)


- Brooke

Robert Hoare

unread,
Dec 17, 2013, 10:02:30 PM12/17/13
to root...@googlegroups.com
On Thursday, December 12, 2013 2:01:42 PM UTC-7, Justin York wrote:
We loaded a subset of the data from the Atlas of Historical County Boundaries. The license is more restrictive than we would like, but we're okay to use it for a proof of concept as we have now. If we don't get approval then there are some other places to look.

It appears that NHGIS https://www.nhgis.org/ has a sort-of suitable licence.  US county and state boundaries from 1790 to the present.  It doesn't give the actual date the boundaries changed, it's just a snapshot every ten years or so, but a good start, at least you then know about the change, by comparing years, and can research the actual date it happened.  They say "all persons are granted a limited license to use this documentation and the accompanying data, subject to [attribution]". 

What are your plans with openplacenames? How will it be different than our project?

I'm focused more on the names and codes from multiple sources, and the relationships between them, rather than GIS.  Some other things as well, best explained by actually putting it online, which I'm trying to do soon.

BTW, you mentioned Worcestershire (England) in your recent video.  It should be pronounced "wooster" not "warsester". :-)

Rob

Justin York

unread,
Dec 17, 2013, 10:52:33 PM12/17/13
to root...@googlegroups.com
NHGIS is a nice data set, though it appears to be more targeted towards enumeration districts so it would be more work to process than the atlas from Newberry. If we can't acquire the Newberry data then NHGIS will probably be our next step.

I'm focused more on the names and codes from multiple sources, and the relationships between them

What do you mean by "codes"? 

--

Tony Proctor

unread,
Dec 20, 2013, 8:30:08 AM12/20/13
to root...@googlegroups.com
I mentioned http://resources.arcgis.com/en/home/ under a different topic in this group since their API seems to cover places and coordinates. Anyone know anything about it?
 
Also, if you want a bit of light reading, I mentioned a few of you all recently at: http://parallax-viewpoint.blogspot.com/2013/12/place-names-or-coordinates.html  :-)
 
    Tony Proctor
Reply all
Reply to author
Forward
0 new messages