Entire moun­tain ranges of data are growing all around, and they will either bury us or help us climb to new heights of under­standing. It all depends on how we respond.

This was the focus of a four- hour “hackathon” Wednesday night, spon­sored by North­eastern to explore the inter­sec­tion between public policy and Big Data analysis. The event, “Data Sci­ence, Jour­nalism, and the Future of Jus­tice,” was part of HUB­week, a series of more than 100 events that brought together the brightest minds in gov­ern­ment, pri­vate industry, and acad­emia to cel­e­brate inno­va­tion in Boston.

“Our goal was to explore the con­ver­gence of data sci­ence in a variety of fields,” said assis­tant pro­fessor of jour­nalism John Wihbey, who hosted the event. He noted that the three pan­elists spe­cialize in rad­i­cally dif­ferent dis­ci­plines, yet all see Big Data as essen­tial in their field.

“In the old days, all you needed was three exam­ples and a quote, and you could report a trend,” said Todd Wal­lack of The Boston Globe Spot­light team and a two- time Pulitzer Prize nom­inee. “But today’s readers expect you to back up your sto­ries with a thor­ough analysis of hard data.”

As an example, Wal­lack described the Spot­light team’s efforts to com­bine tra­di­tional reporting and Big Data analysis to expand a story about teacher- on- student sexual abuse at pres­ti­gious pri­vate high schools. By the time they were done, it had grown into a mega- story that involved more than 100 pri­vate schools in New England.

“Readers want to know the data,” he said. “They want to know how many and exactly which schools.”

Assis­tant pro­fessor Dan O’Brien, an expert in urban studies, ana­lyzes Big Data to dis­cover trends in urban crime and health in order to help cities develop more effec­tive public policy. Com­puter sci­ence assis­tant pro­fessor Michelle Borkin, a spe­cialist in data visu­al­iza­tion, trans­forms com­plex data analysis into visual pat­terns that help guide the work of astronomers and surgeons.

“We are approaching 40 zettabytes of data,” she said to illus­trate the sheer enor­mity of the data deluge. “That’s 40 with 20 zeros fol­lowing it.”

With this in mind, more than 60 par­tic­i­pants launched into the hackathon itself—a two- hour endeavor to make sense of one of three moun­tains of raw data. Each data set came from the Boston Police Depart­ment and focused on either crime rates, homi­cide data, or police stop- and- frisk incidents.

In the old days, all you needed was three exam­ples and a quote, and you could report a trend. But today’s readers expect you to back up your sto­ries with a thor­ough analysis of hard data.
— Todd Wal­lack of The Boston Globe Spot­light team and a two-​​​​time Pulitzer Prize nominee

A team led by jour­nalism stu­dent Abby Skelton, MA’18, deter­mined that women are less likely to be stopped- and- frisked than men of their same race. They also found that the dif­fer­ences vary greatly based on race. White women are 29 per­cent less likely to be stopped than white men, while that figure is 21 per­cent for black women and just 5 per­cent for His­panic women.

Another group, led by mechan­ical engi­neering alum Aditya Agrawal, E’16, used the same data­base to deter­mine that while the number of white and black males who were stopped was about the same, action was taken against a much lower per­centage of the black males. Their con­clu­sion: Blacks are far more likely to be stopped without ade­quate cause.

But given the two- hour time limit, the point of the exer­cise was the process, not the results.

“This was an oppor­tu­nity for data geeks to meet one another, make con­tacts, and be ener­gized by working in a room full of people who share their inter­ests,” said O’Brien, who counts him­self as a data geek as well. “We want to encourage a col­lab­o­ra­tion between dis­ci­plines and agen­cies. Our goal is to get people excited about using Big Data to gain a deeper under­standing of how cities work.”

Story by Bill Ibelle appeared in news@Northeastern posted on September 30.