When Sep Kamvar moved to Cambridge after a dozen years in San Francisco, he was struck by how different the two cities felt. The inquisitive computer scientist, artist, and entrepreneur began to explore how Cambridge’s layout informed his feelings. He tapped his experience studying the visual representation of data, which provides insights into complex data by communicating key aspects more intuitively, like revealing trends in social networks, summarizing huge scientific data sets, and predicting business opportunities. In short, he began to make data visualization maps.
“Every map makes editorial commentary,” said Kamvar, the LG Career Development Professor of Media Arts and Sciences. “A mapmaker communicates what he believes to be important in a city by what he includes in the map, and what he leaves out.” While roads and buildings are typically included, trees and sidewalks, for example, are often excluded, he said.
The Social Computing group is making 10,000 maps—100 maps of 100 cities, including Boston, Cambridge, and New York City. The maps show a wide range of information, including public transportation efficiency, school route walkability, and estimates of footfall density on sidewalks. “The residents will be able to see this narrative of their city that they haven’t been able to see before,” he said.
Building these maps requires software that can analyze large amounts of data. In addition to recording cities’ shapes, software is shaping cities, according to Kamvar. As he was leaving San Francisco for MIT, Kamvar noticed a boom in food trucks and wondered what was driving the trend. “It was because of Twitter,” he said. Food trucks that weren’t properly licensed could avoid parking in the same place, he said. “They would park in different places every day and tweet out their location. They could now have a following without being caught,” he said.
Kamvar is planning to use data visualization to help cities change in conscious ways. He likens the work to performing acupuncture on a city. An acupuncturist’s goal is to unblock energy paths so the body can heal itself. On the scale of a city, one of the biggest energy blocks is lack of information, said Kamvar. Making the information available “catalyzes the city’s natural, intuitive ability to heal itself,” he said.
For example, when sidewalks are too narrow, people are more anxious because, consciously or not, they’re uncomfortable being close to traffic. One approach is to make a map that shows sidewalk widths and places where street parking utilization is under 50 percent. The map gives residents and city officials missing information: places where sidewalks can be widened, he said. “Visualizations really impact the way people think about the world around them,” he said.
Kamvar is also taking big data and data visualization in an unusual direction: from aggregate to individual. He’s made “experiential” data visualizations that uncover patterns in the data and explore data items individually. “Each individual data item is a very rich story,” said Kamvar. “It’s very important to build tools that enable us to go back and forth seamlessly between micro and macro—between large-scale data analysis and the exploration of the individual stories that make up that data.”