News Brief: Research Cited in Senator's Report-Carnegie Mellon News - Carnegie Mellon University

Saturday, May 28, 2011

News Brief: Research Cited in Senator's Report

A Carnegie Mellon University study that explored new methods for studying regional dialects using the Twitter social networking site showed that the use of slang and jargon can reveal important information regarding an author's identity.
   
This finding has important implications for corporations and the intelligence community. The study nevertheless was one of a number of National Science Foundation-sponsored projects whose merits are questioned in a new report released by Sen. Tom Coburn of Oklahoma that scrutinizes the use of taxpayer money for research.
      
"Recent activities in Iran, Syria, Tunisia, Egypt, and Libya are reminders that social networking can be a powerful force in allowing citizens to communicate and coordinate their activities in hostile environments," Randal E. Bryant, dean of the School of Computer Science, said in response to the report. "Unfortunately, these same mechanisms can also be used by computer hackers, criminals, and terrorists. Whether to promote positive uses or to counter negative uses, it's important to understand the nature of interactions via social networking and how they evolve. Carnegie Mellon University scientists are leaders in this area of inquiry.
      
"By its very nature, research requires that scientists pursue promising ideas even — or especially — when the precise outcomes and applications are unknown," Bryant continued. "The Twitter study, which was sponsored by the U.S. Department of Defense and Google Inc., in addition to the NSF, is but one example.
      
"The key finding was that seemingly meaningless slang and jargon can reveal important properties of the author's identity, a point of interest for both corporations and the intelligence community. This finding was the result of a larger, NSF-sponsored effort in machine learning and social media analysis, which enables computers to automatically identify informative patterns in an ocean of noisy data."
      
The Coburn report noted that a total of $1.4 million in NSF grants supported this research. Those grants support general research in social and biological systems; the Twitter study was just one of more than 30 research papers supported in part by the grants.

Byron Spice