Big data knows that most breakups are likely to happen on Christmas, April Fools’ Day, Mondays and during the summer break.

This and other applications of big data are clues to its possibilities for the future, said Viktor Mayer-Schonberger at a lecture on Wednesday.

Mayer-Schonberger, a professor of Internet governance and regulation at Oxford University, spoke in Goldstein Auditorium. He discussed the benefits and drawbacks of big data, which is the collection of data on a massive scale. He said it can quickly provide updates on the spread of a disease or help doctors analyze trends in that data to predict infections in premature infants. But, its ability to predict human behavior could also lead to the creation of a surveillance society and possible issues with premature criminal punishment, he said.

One of the positive components, he said, can be seen in 2009, when the H1N1 virus was circulating around. The Centers for Disease Control and Prevention in Atlanta was getting readings on the spread of the flu, but the information was often two weeks behind, Mayer-Schonberger said.

Google then used its search engine to narrow down billions of searches down to 46 terms related to the flow of the flu, he said. He explained the CDC was able to use these terms to get up-to-date information about the flu’s spread substantially faster as a result.

Big data has also benefitted premature babies, he said. It’s provided a way for doctors to manage and predict infection in newborns so they could give life-saving treatment.

Mayer-Schonberger also discussed the growth of data and how it was able to improve to the point where it could yield these benefits.

In 20 years, the overall amount of data has multiplied 100 times, he added.

Currently, less than 1 percent of the data that exists is non-digital, while 13 years ago, three-fourths of the data in the world was non-digital, Mayer-Schonberger said.

This growth is ongoing, as digital data becomes the primary way of storing information, he said.

Mayer-Schonberger went on to discuss big data’s potential drawbacks. It could, for example, create a surveillance society, he said.

More and more technology is able to predict people’s behavior by tracking certain web activities, which he called “dangerous.”

Another hypothetical situation, he said, is behavior-predicting algorithms playing a greater role in the criminal justice system. This would predict crime before it happens and giving out punishment without proof, he said.

Bradley Leverence, also an undeclared freshman in the College of Arts and Sciences, said he thought Mayer-Schonberger’s presentation was fascinating.

“I found it shocking how today less than 1 percent of the data that exists in the world is analog while it was three-fourths of the total data little over a decade ago,” he said.

Mayer-Schonberger ended his lecture with a warning to maintain perspective on big data and tread carefully.

Said Mayer-Schonberger: “It is just a shadow of reality and never will be fully accurate. And as we walk in this age, we must do so with humility.”

Published on October 10, 2013 at 2:15 am