Menu Bar

Home           Calendar           Topics          Just Charlestown          About Us

Thursday, March 10, 2016

Coping with “big data”

From writing and rhetoric to oceanography, big data is central to research, education at URI

KINGSTON, R.I. –A recent story in the Telegraph of London says people are “bombarded by the equivalent of 174 newspapers of data a day.” Another story in the Telegraph uses the word vlogger, which the e-communications world describes as a video blogger.

And who is trying make sense of all that and preparing students to be effective writers, editors, producers and web strategists in this dizzying world of instant, multifaceted communication? Scholars in the University of Rhode Island’s Department of Writing and Rhetoric. And like their colleagues in science, engineering, mathematics and computer science they are part of the Big Data Initiative at URI.

Even writing and rhetoric URI doctoral student Jenna Morton-Aiken is using big data to organize 300 boxes containing thousands of written records for URI’s National Archives of Composition and Rhetoric.

“We are digitizing the records and organizing them in a framework that will allow future users and contributors to interact with the data, and interact with each other,” Morton-Aiken said. “We are going from a hierarchical, physical collection of records to a network using hashtags. People from anywhere in the world will be able to examine the last 30 years of writing teachers’ archives on the topic of writing and composition to see what has worked and what hasn’t and what they might try next.”

Associate Professor Jeremiah Dyehouse, chair of the Department of Writing and Rhetoric, said that an “Infopocalypse” has developed in the last 50 years, during which writing, or more appropriately, communicating, is no longer linear.

“For instance, old writing had a beginning, middle and end. That included books, essays, newspaper stories and even recipes. But with the Internet and social media, even those old media forms are subject to online comment and are supplemented with video and photos often provided by those who are not the original authors,” Dyehouse said.

And not only is social media and the web affecting old media, there is original content being created online in blogs, videos, Twitter, Facebook, Snapchat and Instagram, among others, that will never be seen on the printed page or on a traditional television station.

“All of this changes the way we teach,” Dyehouse said. “We are looking at how can we wade through the ‘Infopocalypse’ and teach our students and society at large so we can have an informed and engaged citizenry. Jenna’s work is a concrete example of how rhetorical scholars can use big data to continue learning in this complex environment.”

Big Data on the oceans

Meanwhile, just a few miles away at URI’s Graduate School of Oceanography at the Narragansett Bay Campus, Professor Peter Cornillon is using an algorithm that allows him to calculate sea surface temperatures using global satellite data.

Cornillon calls himself a satellite oceanographer who studies the physics of the oceans, with particular attention paid to sea surface temperature fronts. Cornillon has more than 500,000 sea surface temperature fields from more than a dozen satellites, covering the global oceans for the past 30 years. These data total about 100 terabytes. A terabyte is about 1 trillion bytes, or 1,000 gigabytes.

“By contrast, the first disc drive I bought cost me $20,000 and it held 400 megabytes, which in 1982 was considered big data,” Cornillon said. “These data are about 250,000 times greater than what was on that first disc.”

All of the data Cornillon works with is in the public domain, which means scientists from all over the world in numerous disciplines have access to it. 

He is also working with a scientist at NASA’s Jet Propulsion Laboratory who has just run a global ocean model covering a two-year period, producing about 1.5 pedabytes of output--15 times the size of Cornillon’s data.

“This is a complete model providing very high resolution near the (ocean) surface,” he said. “I would like to include in my courses access to this model run allowing students to take digital cruises to collect data while on digital cruises, thus simulating what one would collect while on a ship,” Cornillon said. 

Cornillon’s colleague at GSO, Isaac Ginis, an internationally renowned expert on hurricanes, said his work involves sophisticated modeling that requires high performance computing, but almost all of his computing is done at the U.S. Navy, National Oceanic and Atmospheric Administration and other universities. That group includes the University of North Carolina at Chapel Hill, which like the Cybersecurity and Digital Forensics Center in the Computer Science and Statistics Department at URI, has been named a national Center of Excellence by the U.S. Department of Homeland Security to research coastal resilience.

“We are developing high resolution simulations of high hurricane impact areas in the face of climate change, and it will be a great advantage to our students to have access to theses simulations right here at URI,” Ginis said. “We could attract more undergraduate and graduate students to learn how to apply these tools to weather and climate questions. Such capability would make us competitive with other universities and agencies, and we would be able to collaborate more fully.”

Big Data in pharmacy

That’s also true in health care where Pharmacy Professor Stephen Kogut is using big data from private health insurers and the state’s Medicaid systems to analyze diseases in populations, cost, medication use and hospitalizations. For Rhode Island Medicaid, Kogut and his team analyzed more than 2 million pharmacy dispensing records to try ascertain medication patterns for those with depression to see who might be continuing treatment as recommended.

“Data sources for these areas are common in health care outcomes research,” Kogut said. “We are looking at payment for insurance claims. Insurance companies use the data for their own internal needs for setting premiums, but they provide us with information about outcomes and costs. We get great data that can lead to interesting and important research projects.”

In addition to data from insurance companies, internet data, patient records, doctors, hospitals, imaging and blood laboratories, Kogut said there is a new and growing source—smart devices.

“These devices can track your pulse, sleep patterns and test your blood,” Kogut said. “Imagine if an hour before you have a heart attack, your smart phone sends a message to your doctor based on the data it is collecting from your body.”