SUNY Oneonta of the Future: Big Data U
By JIM KEVLIN
Last year, sociology professor Brian Lowe’s “Animals and Society” class read Jon Mooallem’s “Wild Ones,” where the author draws conclusions about how humans are coming to grips with the extinction of wild animals, polar bears, a type of butterfly and whooping cranes in particular.
“Its central theme,” as the review in Mother Jones had it, “has more to do with the qualities we project onto animals than the creatures themselves.”
Most classes would require undergrads to understand how Mooallem came to his conclusions, but entering senior Matt Hartwell took it a step further, seeking objective proof on the validity of the book’s conclusions.
Using IBM’s Modeler tool, Matt was able to sort through the world of Twitter. He devised key words – names of endangered animals and opinion leaders – and applied them to several thousand “tweets,” using the University of Buffalo’s Supercomputers to come up with his results.
Hartwell was able to conclude that, among Twitter users anyhow, not polar bears or whooping cranes but elephants and rhinos are the threatened animals humans care about most. And the “opinion leaders” with the most influence in the matter were British royalty – Prince Charles and his son William – and soccer star David Beckham.
This fall, Achim Koeddermann, philosophy professor – “I would have voted on apes,” he said – will be taking those findings to his environmental ethics class. “Why were these animals chosen? Do they look like us?” he asked. Are we channeling “Babar The King?” Is it habit? Culture? Age?
If this scenario sounds unfamiliar, it should. Only at SUNY Oneonta – to repeat, ONLY at SUNY Oneonta – are these kinds of Big Data experiments being conducted at the undergraduate level. And nowhere, undergraduate or beyond, would you find a sociologist, a philosopher, various information technologists and a student freely sharing their opinions and conclusions.
And they were doing so about what is arguably the dominant and most debated post-Edward Snowden issue of American life: Big Data.
SUNY Oneonta’s Big Data program, Lowe said, grew out of admiration for statistician Nate Silver, whose New York Times blog, fivethirtyeight.com, correctly predicted the outcome of the Obama-Romney 2012 election in all 50 states and the District of Columbia, and 31 of 33 gubernatorial races.
In addition to Lowe and Koeddermann, the original team included Jim Greenberg, director of the Teaching, Learning and Technology Center; sociologist Greg Fulkerson, and Bill Wilkerson and Brett Heindl in political science. With the help of Kathy Meeker, the campus’ chief grantsperson, a $20,000 SUNY Innovative Instruction Technology Grants in 2012 moved the effort forward.
Eight active participants sat down the other day for an hour-long conversation with a reporter. In addition to Lowe, Koeddermann and Hartwell, there were Tim Ploss, Diana Moseman and Chilson Reynolds from IT, and chemistry professor emeritus Harry Pence, who remembers the celebration when the campus’ total computer memory rose from one to two gigabytes, and he’s stayed in touch with the evolving scene ever since.
In the beginning, they related, conversations about an affiliation were opened with Penn State, but the collaboration with SUNY Buffalo’s Center for Computational Research and its powerful computers soon emerged as the better option.
The collaboration “demonstrates the power of SUNY,” said Pence. Only within SUNY – so far, anyhow – has an undergraduate campus been able to gain access to the computer firepower necessary to process “the whole fire hose” of data, as Moseman put it.
Practical applications are already evident.
After the Matteson Hall dorm caught fire in 2013, the Big Data team was able to identify three possible causes – a candle, a space heater or a refrigerator – all banned in dorm rooms. Alerted to these conclusions, SUNY police waived potential penalties in interviewing students: More important than blame, they concluded, was simply finding out what happened.
Fulkerson is using data to explore what people, unfiltered, actually think about fracking, a matter of local interest, which he aims to include in a book.
Still, the team acknowledges, it’s early in the Big Data game. For instance, Twitter tilts young and educated. The “firehose” will have to be expanded to include Facebook, e-mails and beyond, and even then will be defined by the opinions of the digital savvy.
But it’s always been that way, Pence and Koeddermann said: The well-respected Pew surveys, for instances, have tilted toward people with landlines, increasingly less relevant as time goes on.
Another area of issues is ethical, dramatized by the NSA’s much debate spying on Americans. “Privacy is dead,” Pence says bluntly. “The floodgates are open,” said Lowe. In the Matteson investigation, for instance, identifying data was removed from the tweets, yet students who saw them recognized the authors.
The difference in the SUNY Oneonta effort, Koedderman believes, is that sociologists – and, in his case, philosophers – are the drivers. “Critical thinking will still be the backbone,” he said.
Where all this is going is unclear, but it is clear that Big Data is probably only going to get bigger. For now, Big Data techniques are being incorporated into existing courses, but Lowe hopes before long a Big Data 101 becomes plausible.
Already, one of Lowe’s students gained a post-graduate berth at Rice University based on his Big Data experience here.
You can see how such a unique offering would help draw talented students to SUNY Oneonta. So far, Big Data hasn’t been promoted in that way. But maybe it’s time, the gathering agreed.