Data is out there. I would refrain from saying good data is out there, or even relevant data is out there. As Kim said we are in the “year of the infographic” we are equally in the “year of the unruly excel document”. When one is lucky to receive large data that is relevant, they may stare down at their excel or SPS document and then say: “Now what?”
One of the types of data become more and more accessible is addresses. More and more companies and organizations post their address online. They are aggregated up into somewhat tidy databases like RefUSA and OneSource, accessible on many college campuses and local public libraries. Once bound in the white pages, this address data is the meat of many new and exciting ways to research. With information about where companies are, you can find a great number of exciting things. For example, the location of food stores in Detroit can help you locate food deserts. The location of Targets in the United States can help you predict where next to put your big box store. I just had a chat yesterday with a professor who works how industry is affected by nature disasters, using location-based disaster data and addresses as well as other indicators.
Sold on address data? Excellent! We’re going to go over some tools for address data. Firstly, if all you want to do is take those addresses and put them on a map, BatchGeo is an excellent tool for taking all those address and putting them on a google map. You can code 150,000 address per IP per day, and creates a google map which you can then improve on the google interface, either drawing polygons or adding metadata. But what if you want to do some other visualizations?
Google Fusion Tables is an awesome product. Google Fusion Tables also can geocode addresses into a google map.
Why I like Fusion Tables:
• Different types of visualization
• Sweet, sweet fusion. The merging capabilities of Google Fusion
• Collaboration. You can share your data sets relatively securely via Google Fusion Tables with research partners.
• You can link it up to Google Refine for those more squirrelly datasets.
Here’s an example with some roughly 300 gas stations in the city of Detroit.
Here’s the data as it looks in excel. Ugly.
Here it is geocoded:
Another view (this time using some sales data as well)
Another view (more pie charty). I understand this is a terrible pie chart, but it’s very aesthetically pleasing.
For a relatively boring but very effective run-through what makes Google Fusion Tables wonderful, see below.