Morarescu Speaks About Data Use, Analysis on Big Data Day Panel
By: Diane Stirling
Assistant Professor Paul Morarescu (right) on the panel at Big Data Day.
School of Information Studies (iSchool) Assistant Professor Paul C. Morarescu was part of a panel on Monday night for Big Data Day at Syracuse University, presenting his views on the use of data analytics in the last presidential election, big data’s place in future elections, and how data analysis will increasingly impact politics, business, and most other aspects of American life in future years.
The discussion was hosted by the “Navigate New Media” group at the S.I. Newhouse School of Public Communications. Joining Morarescu were Grant Reeher, professor of political science and director of the Alan K. Campbell Public Affairs Institute in the SU Maxwell School of Citizenship and Public Affairs; and special guest Ashley Bryant, the state digital director for Ohio for the Obama Campaign in 2012. The panel was moderated by Hub Brown, an associate dean at the Newhouse School of Public Communications.
Morarescu joined the iSchool last fall. He teaches courses in the school’s Certificate of Advanced Studies program in Data Science. His research is focused on natural language processing, question answering, information retrieval, information extraction, and big data.
The use of data analysis to drill down to mine information about people is a practice Morarescu said he is “enthusiastic about, but I’m also trying to be objective and trying to think about the implications of using more data as opposed to the human factors, as there are privacy concerns,” he said.
Morarescu told the audience that as big data matures as a science, its use will become more prominent and will impact “every aspect of our lives, not only politics or university classes and curriculums, but every sector of every industry. The financial sector is impacted so much already and big data will impact other industries as well,” he noted. Citing statistics showing that one-third of business executives currently make decisions using data that is missing or that they don’t fully trust, he predicted that business and industry will make big investments in the future in order to expand data analytics as a resource.
The increasing sophistication of data analytics is highly visible in the political sphere, Morarescu said. The Obama campaign, noted for its sophisticated use of voter data in grassroots organizing and support efforts, showed vast differences in data practices between the elections of 2008 and 2012, he said. That included a five-time expansion of its data analytics team in the latter campaign. According to Morarescu, the 2008 campaign used big data to determine only the current status of voters and if they were likely or unlikely to vote. Four years later, new dimensions in data practices were added to gauge predictability of voter behavior. “It added a whole new dimension, from just monitoring the status of the voter to trying to change the behavior of the voter,” he reported. “My guess is that is going to keep growing and the importance of data in elections will keep growing as well.”
The professor also described the kinds of skills and tools that students seeking careers as data scientists and engineers should possess. They are: “statistics, for data mining and looking for patterns; computer science, for how to program computers and R language; and you need to be a good communicator, because you’ll be communicating your results to a wider audience.”
The field has changed in that “a data scientist no longer sits at his computer and only communicates with his team and his manager,” Morarescu explained. “You need to be able to communicate to top management of your company, and to politicians, voters, and others. And if you go into business, you need to know something about behavior, economics, and social psychology. You need to be kind of a field expert so you know what kind of data to put into your models to get results,” he said.
While the discipline is so new that “colleges in the U.S. are still working out courses and curriculums,” big data is “more and more a part of our lives, and more and more intrusive; there are many questions about it and many more issues we don’t understand yet,” he added.
“Many, many online marketing companies already have a lot of information and they track people online. A basic understanding of one’s privacy rights online is not sufficient to address the entire issue, Morarescu said. Even though there are legal people and policy experts looking for solutions to the online privacy issue, it is a situation where “I don’t see a solution anytime soon,” he concluded.
Before joining the iSchool, Morarescu was a research computer scientist in the Speech Technology and Research (STAR) Lab at SRI International in Menlo Park, CA; a principal software engineer for Ask.com in Edison, NJ; and a postdoctoral researcher for the IBM T.J. Watson Research Center in Hawthorne, NY. He earned his Ph.D. in Computer Science from the University of Texas at Dallas in 2007.