GAFFTA visualizing and mapping data class, day 1

By Marcus Wohlsen | Feb. 9, 2012, 4:17 p.m.

I had a fantastic time at the Gray Area Foundation for the Arts' visualizing and mapping data class, which met for the first time Tuesday. The first night we built a rough approximation of Google Maps using open-source mapping data and a little Javascript. You couldn't search _ obviously a key feature _ but you could pan and zoom your way around the whole world.

The key elements were a map-centric Javascript library called Modest Maps and the maps themselves, which came from a service called Cloudmade. Onto this map we placed historical photos of San Francisco shot around the time of the 1906 quake at the spots where they were taken. These were pulled from an online collection of historical photos called Historypin. All these resources, of course, live in the cloud. Our html file that tied them all together came in at ten kilobytes.

The simplicity of the project may not surprise designers who work with this kind of data a lot, but as a journalist still in the early stages of exploring these technologies, I was gobsmacked. I came into the class excited about data visualization as a medium for telling stories in new ways. I came away from the class thrilled at the breadth of opportunities to do cool stuff in that medium even for those of us with modest coding backgrounds.

The class is taught by Michal Migurski, director of technology at San Francisco-based Stamen Design, which specializes in map-related projects. We spent the first part of the class looking some of Stamen's work and projects by others using maps and data in innovative ways.

One of the apparently best-known examples to merge social media data with maps is Eric Fisher's "Locals and Toursits" map of San Francisco. Fisher created an algorithm that analyzes the Flickr photo-posting habits of users who've shared pictures geo-tagged to specific locations. Based on how often they post photos from specific locations, he identifies each user as a "local" or a tourist. He then plotted the pictures taken by each on a map of the city (locals are blue, toursits are red, yellow is either):

Locals and Tourists #3 (GTWA #4): San Francisco

Making use of city data easy to come by thanks to San Francisco's open data initiative, Shawn Allen created a San Francisco map that plots cabs versus crime versus trees:

Trees, cabs and crime in San Francisco

I consider both of these projects journalism. They each take raw, relatively obscure data, organize it and present it in an engaging, user-friendly way that tells a story.

Stamen itself has created projects that are even more plainly journalistc, such as their Oakland Crimespotting map. Migurski said crime data is often the first thing newbies reach for when working on map visualization projects because it's ubiquitous and relatable. Compared to what local police departments offer, the Crimespotting maps are an order of magnitude more flexible and more usable.

For MSNBC, Stamen built a storm tracker that pulls in hurricane data to create maps that not only track storms but give residents of threatened areas a graphically graphical sense of whether their communites are in a storm's path.

Especially with these last two, it's easy to see how you could layer more traditional journalistic elements over these maps, such as video, photos and text, to create an even richer narrative experience. I can envision a future of journalistic apps that knock down a barrier I think we journalists often unconsciously try to maintain between story and utility. I love the idea of interactives that both tell stories and serve as tools. The possibilities seem endless, and from what I'm seeing so far in this class, well within reach.

blog comments powered by Disqus