GeoCommons Part of the NSDI!?!? Making Tabular Data Magically Become GeoData

Chalk one up in the irony column. We applied for an FGDC NSDI CAP grant and actually won. For those not familiar with the alphabet soup FGDC is the Federal Geographic Data Committee, NSDI is the National Spatial Data Infrastructure, and CAP is the Cooperative Agreement Program. While I’ve voiced my opinions on the current state of SDI’s and using bailout money to fund it on this blog, I do believe the goal of this grant very much points us in the right direction.

Specifically the grant we applied for was “Enabling Use of Government Tabular Data in a Geographic Context”. I believe creating this grant category showed a lot of foresight from the FGDC. I don’t have insight on what their rationale was, but I can say what got us very excited about it. Primarily this is a concrete step towards making geographic location a mainstream part of all data, and potentially a step away from geographic data being a special niche separate from the rest of data. Mapinfo and Gartner had a statistic a while ago that there was a geographic component to 80% of tabular data. By my calculations that means you should have the option to view any data geospatially and only 20% will be missing out.

When we start thinking of geo being a part of all data it really starts to open up the door to bringing GIS and neogeography fully main stream. Imagine if all the information in the speculated Data.gov had the option of consuming it geospatially. If projects like this NSDI grant are successful the ability to intelligently assign location (point and polygon) to all relevant tabular data provides the opportunity to make geography a part of everyone’s work flow.

In the “leg up” category we’ve been working and prototyping around this problem for a while, and have a pretty good head start. There is still plenty of work to do but, I thought this would be a good opportunity to get feedback from the community on our approach. If there are formats you would like to see supported or functionalities that would really extend the capability please pass them along. Below is a diagram that should provide a high level perspective of what we have in mind at this stage: