By Rob Mitchum // May 2, 2013
Lost in the buzz surrounding “Big Data” is the nuance that the phrase can mean very different things according to its context. In the sciences, “Big Data” is a matter of increasing scale for researchers working with massive datasets in physics, astronomy and genomics. In business, “Big Data” offers new ways of marketing to customers or engineering better products and projects. But for a broad segment of the population, “Big Data” has yet to directly impact their daily lives beyond targeted ads on the internet and personalized recommendations for books, music and movies.
In the Computation Institute’s April 25th panel for the international Big Data Week event, three panelists highlighted the potential of this fashionable phrase to make the world a better place. Where data analytics and high-performance computing have long been essential tools in the physical sciences — and more recently in biology and medicine — the crossover of these methods into the social sciences is just starting. Now, as government data becomes more accessible, either publicly or for research purposes, new opportunities to improve the world around us arise, from building more sustainable and healthier cities to protecting society’s most vulnerable citizens from harm.
Charlie Catlett, director of the CI’s Urban Center for Computation and Data, started off the session by urging attendees to “think exponentially” about the world’s problems. Today’s smartphones hold as much computational power as the most advanced, room-sized supercomputers of 25 years ago, Catlett said, illustrating how quickly technology evolves. To maximize the potential of science along this exponential curve, researchers need to aim at problems that may seem insurmountable at present.
“When we’re thinking of a problem or a solution, we’re often overly constrained by what we understand and what we know and what we can do today,” Catlett said. “Where the real creativities and breakthroughs come is when we say, ‘I know I can only do this today, but I also know in five or ten years that the technology will catch up with that vision I have.'”
For Catlett and UrbanCCD, that means arming governments and developers with the computational tools to build and expand cities at unprecedented scale. Over the next decade, China will have to construct city infrastructure to house 400 million people — the equivalent, Catlett said, of 10 New York Cities. New, powerful computer models feeding off actual city data are needed to predict the effect of these new megalopolises on the environment, on services such as energy, healthcare and transportation, and on the city residents themselves. As a test run closer to home, the UrbanCCD is working with Skidmore, Owings and Merrill on the Chicago Lakeside project, a 600-acre development on the city’s South Side.
“Unlike most other cities in the developed world, we have an opportunity to work with developers and the city on a project that’s of a scale relevant to the rest of the world,” Catlett said.
At the other extreme, Bob Goerge of Chapin Hall talked about his work using public sector data to identify children and families at risk of abuse, incarceration, substance abuse and other negative outcomes. For 30 years, long before “Big Data” was a topic of discussion, Goerge has gathered and combined data from various government agencies to find individuals and families that use multiple social services, such as Medicaid, food stamps, unemployment assistance or mental health treatment. With a database containing millions of individuals, the researchers could identify families and neighborhoods at higher risk for violence, incarceration and other dangers — knowledge that could help governments more effectively target their resources and interventions.
To continue such important work and launch even more ambitious data projects will require creating an entirely new culture of socially important data science, the subject of the panel’s final talk from Rayid Ghani. Ghani told his own story of moving from business data analytics for Accenture (who hosted the panel at their Chicago headquarters) to President Obama’s re-election campaign to his new home at the CI, searching for a career that could unite his passions for volunteering and computer science. This summer, he will lead an effort to train nearly 40 more scientists with similar interests during the inaugural Data Science for Social Good fellowship. The fellows will spend 12 weeks attacking data-intensive problems in transportation, health care, public safety and more, which Ghani hopes will lure them away from more fashionable tech jobs to a career with lasting social impact.
“The goal is to not just solve these problems, because the solution will often be long term,” Ghani said. “Part of the goal is to get these students really talking about and excited about these problems so that a handful of them hopefully change their minds about what they’re going to do next.”