Jeannette Wing on Developing 'Data for Good'

May 04, 2018
Jeannette Wing in a beige suit, in a white, well-lit room

From filter bubbles to fake news, data-driven algorithms have developed a reputation problem. Jeannette Wing wants to change that. As the new director of Columbia’s Data Science Institute, she has woven ethics and social impact into the Institute’s mission. Her vision comes amid a backlash over the tech industry’s use and misuse of consumer data and a broader shift. The exponential growth of digital data, and tools to harness its power, have disrupted much of society, from our economy to our politics.

An accomplished computer scientist, Wing sees a critical leadership opportunity for Columbia. The Data Science Institute was founded six years ago at the Engineering School, but with her arrival last summer, it became a stand-alone entity. From this elevated platform, Wing has called on social scientists and humanists to work with data scientists to ensure that data is used responsibly and for society’s benefit.

“Who best to help, but people who think about ethics and social norms?” she says. The campus has embraced her challenge.

“The Data Science Institute had to begin in Engineering, but to realize its full potential it couldn’t end there,” said Matthew Connelly, a history professor and Institute member. “Data science will eventually be a part of education and research across the entire academy. All the more reason we need to consider ethics and how data science will shape, and be shaped by, history, law, public policy, and so on.”

In an uncluttered office with floor to-ceiling views of northern Manhattan, Wing sat down recently to review her first year at Columbia.

Writing longhand in an old-fashioned notebook, she sketched out her plan for building “a cadre of data scientists” to meet the rising demand on campus. She has recruited three types: applied data scientists to serve the University at large, domainspecific faculty, and postdoctoral researchers to bridge data science and a domain.

She has also launched a summer internship program that matches data science students with faculty research projects. In addition, she has doubled the Institute’s seed funding for high-risk projects. One such gamble pairs a computational linguist with an astronomer in a hunt for “grammatical” rules that explain how planets form.

Her unconventional approach is reflected in her enthusiasm for other universities opening competing data science institutes. This spring, she invited 60 of her peers to Columbia to share best practices. “I want people to look to us as a leader in data science,” she said.

Today, the field remains as poorly defined as computer science was in the 1970s when Wing was studying at MIT. She had planned to follow in the footsteps of her father, an electrical engineering professor and department chair at Columbia, but a required computer science class seized her imagination. After finishing her bachelor’s, master’s and doctoral degrees in computer science at MIT, she eventually joined Carnegie Mellon University as a professor and later chair of its renowned computer science department.

Later, she held executive positions at the National Science Foundation and Microsoft Research where she became known for embracing audacious ideas. At Microsoft, she launched an Expeditions Program modeled after the one that she created at NSF. It led to a new application for drones as a cropmonitoring platform and a smart mosquito trap that uses machine learning to identify species by the flap of their wings to more quickly detect the spread of Zika and other mosquito-borne diseases.

In an influential 2006 essay, “Computational Thinking,” Wing argued that the abstract thinking required in computer science is fundamental to all problem-solving. Much of her own research has focused on cybersecurity and privacy, including work automatically generating examples of potential cyberattacks and building scalable tools to detect violations of privacy policies.

As a leader, Wing has had unusual success getting creative, independent thinkers to work as a team. One of her secrets was revealed at a faculty retreat last fall when she proposed a game of charades. “For a few seconds I thought she was joking,” says Chris Wiggins, an applied mathematics professor and Institute member. “But then I thought, OK! I’m about to learn a lot about Jeannette’s collaborative style.”

By the following day, the group had agreed on a definition for data science: the study of extracting value from data. They had also drawn up a list of socially relevant “quests” to tackle—from detecting the spread of disinformation online to motivating healthy habits by merging personal, genomic, and environmental data to helping cities adapt to a warming climate. Wing’s hope is that solutions to complex problems can be found if experts across disciplines identify and rally around a shared goal that can benefit society. 

This ability to listen and let others lead has won over colleagues. “She’s an active listener who provides constructive comments,” said Garud Iyengar, chair of Columbia Engineering’s Department of Industrial Engineering and Operations Research and the Institute’s deputy research director. “I walk away thinking that the idea I came to her with got improved upon because of her experience with technology, people, and problem-solving.”

Wing arrived at Columbia as the College of Dental Medicine was preparing to launch its Center for Precision Dentistry. Dean Christian Stohler says Wing’s support was key in convincing skeptics that a dental clinic outfitted with sensors and cameras to measure every aspect of the patient experience would be worth it.

Her involvement didn’t stop there. At a recent talk at the school, Wing opened a discussion on data privacy, prompting several students to share their concern about cameras recording their every move, including their mistakes. Stohler reassured them: “The data are for the benefit of students. We are not selling the data to Russia! We are keeping it here.”

Big data’s promise and perils were on display. So were the challenges of ensuring that data will be used responsibly, in perpetuity. Solutions are never easy, says Wing, but starting the discussion is a first step—one of many more to come.

“I feel I can’t move fast enough but I also know I need to be patient,” she said, before citing the title of a favorite Piet Hein poem. “Things take time.

—By Kim Martineau