Project compiles Instagram posts from different cities to explore patterns in the data and create visual signatures for each city.
When there is a beautiful pink sunset or a massive moon, chances are you’re not the only one Instagramming it. Such is the world we live in today, that many experiences are complete only when shared with others. While researchers have been working hard studying social media data from the likes of Twitter, Facebook and Foursquare, the visual data from Instagram has proven a much harder nut to crack. But a team from the Art History department at the University of Pittsburgh, the Software Studies Initiative at California Institute for Telecommunication and Information and the Computer Science program at The Graduate Center, City University of New York have come together to study 13 cities and the Instagrammers therein, to examine what personal photos can tell us about the city as a whole.
The project, Phototrails consists of city visualizations from places as diverse as Beijing to San Francisco, created by aggregating users photographs and measuring different visual attributes such as hue, brightness, line orientation. The result was a series of unique ‘visual signatures’ for each city. For instance, by compiling photographs and organizing them based on color and timestamp, one can see the daily rhythms of a city (Tokyo below) as the hues shift from light to dark – daytime photos to ones taken after the sun sets.
The visualization below shows the shift in users interaction with Instagram during Hurricane Sandy. Prior to the power outage there is a plethora of photos documenting the incoming storm, and the outage itself is clearly demarcated where the pattern becomes less busy and there are much darker photographs being taken.
Lev Manovich, a main collaborator on the project, explained to Fast Co.Create:
We see a clear line that separates ‘before’ and ‘after’ the event. This sudden and dramatic visual change reflects the intensity of the experience.
The Phototrails study captures a part of social media that has thus far been understudied as a form of Big Data. Next Manovich and his collaborators Jay Chow and Nadav Hochman will be moving onto Facebook to visualize our daily photosharing on that medium and what it may indicate about wider cultural behavior.