IU research on social media attacked as U.S.-funded Big Brot
Source: Maureen Groppe
Can tweets predict the outcome of elections and shifts in the stock market?
How can you tell whether a tweet was sent by a human or by a computer pretending to be a human to sway public opinion?
Does the spread of social media really mean more people can have influence, or does power still rest with a select few?
Those are some of the questions researchers at Indiana University’s School of Informatics and Computing have been trying to answer as they study how memes ― ideas, values or patterns of behavior ― spread online.
The work has been supported with about $1 million in federal funding from the National Science Foundation since 2011. Researchers have published their findings in about 30 papers in peer-reviewed journals and conferences. They have received extensive media coverage. And they have even consulted for an episode of the CBS television show “The Good Wife,” in which the main character represents a client incorrectly identified as a terrorist on a social website.
But in recent months, the “Truthy” project also has been lambasted in social media, by conservative news organizations and by some public officials as a government attempt to monitor public opinion and track hate speech. The criticisms culminated last week when the GOP chairman of the House Science, Space and Technology Committee announced that his panel is investigating how IU’s research project received federal funding.
“While the science committee has recently looked into a number of other questionable NSF grants, this one appears to be worse than a simple misuse of public funds,” Rep. Lamar Smith, R-Texas, said in a statement. “The NSF is out of touch and out of control.”
The National Science Foundation said through a spokeswoman that the agency “stands by its rigorous merit review process.”
Most of the thousands of grant proposals the agency receives are evaluated by at least three independent reviewers consisting of scientists, engineers and educators who are experts in their field, said spokeswoman Dana Topousis.
Indiana University likewise issued a statement of support for the researchers and staff, who have received hate mail.
Filippo Menczer, the director of IU’s Center for Complex Networks and Systems Research, where the work is being conducted, said it’s ironic the project is the subject of the same type of misinformation campaign that it studies.
“Whoever is pushing this misinformation is able to get other news venues to believe it or repost it or republish it,” Menczer said. “That’s the mark of a very successful meme.”
Critics describe the project as Big Brother, tracking what people are saying on social media.
“Now you’ve got another federal agency that is going to monitor you and your misinformation,” Fox News anchor Megyn Kelly said in an August broadcast. “How is that going to be used against the average Joe?”
A member of the Federal Communications Commission asked in an opinion piece published by The Washington Post last week why the government “is spending so much money on the study of your Twitter habits.”
“If you tweet your support for a candidate in the November elections, should taxpayer money be used to monitor your speech and evaluate your ‘partisanship’?” asked Ajit Pai, one of the FCC’s five commissioners.
Smith, the chairman of the science committee, said the government has “no business using taxpayer dollars to support limiting free speech on Twitter and other social media.”
Menczer said the critics’ descriptions of IU’s research are false. They take parts of the work out of context to make it sound sinister.
Neither IU nor the government is controlling, or even monitoring, tweets of individuals. And researchers are certainly not evaluating the “correctness” of every tweet, he said.
“We don’t sit there and say, ‘This is true. This is not.’ It would be crazy,” he said. “We have millions of tweets a day.”
What they are doing is aggregating millions of public tweets and using complex computer models to look for behavioral patterns. Anything they find is made public, along with their data and tools.
Their first research projects focused on the 2010 elections, because it was easy to extract relevant information from tweets, such as the names of candidates. The papers they published included how to tell whether a social media trend is being generated by computers or is organic. And they looked at how much interaction there was among Twitter users of different political perspectives.
Other research projects have analyzed Twitter’s role in the Occupy Wall Street movement and in the demonstrations and civil unrest in Turkey last year.
As for tracking hate speech, IU said in its NSF grant proposal that it would create a publicly available site to monitor trends in social media that could “mitigate the diffusion of false and misleading ideas, detect hate speech and subversive propaganda, and assist in the preservation of open debate.”
The website already includes a “bot or not” demonstration where anyone can type in the name of a Twitter handle and get an estimation of whether the poster is a robotic computer program or a human.
That tool stems from work for the Defense Department. Researchers got about $2 million in 2012 to figure out how to detect “persuasion campaigns” in social media in the early stages, before they become indistinguishable from spontaneous methods of idea sharing.
While Menczer emphasizes that he and his fellow researchers are not policing the Internet, he said the work they’re doing to help people identify misinformation campaigns can help preserve free speech.
“If your opinions are based on things that are factually false, that hinders free speech because you have to be informed to have an engaged and useful conversation,” he said.
Being able to spot misinformation campaigns also can stop people from being duped into buying junk stock, or from panic because of fake reports about the Ebola virus spreading through an entire classroom because of an exchange student ― as was claimed by a fake news site.
Examining how people get and share information are fundamental questions scientists have been asking for years, Menczer said, and social media provided the massive amounts of public data that researchers can use to find answers. Plus, social media platforms play an ever-increasing role in the lives of people.
“We use them to communicate with other people, whether they’re friends or family or celebrities or people we don’t know. What we see and what we read affects what we know and what we think and what we believe and how we behave,” he said. “Whether we’re aware of it or not, we are affected.”
The project’s name, however, is not helping IU fight its critics.
Fox News said the “bent” of the researchers could be gleaned from the fact that “Truthy” was inspired by comedian Stephen Colbert’s term “truthiness,” which describes claims that sound true but aren’t.
Menczer said a graduate student suggested the name, which struck others as cool and funny.
“And so it stuck,” he said. “Never would we have thought back then that, years later, this would somehow make people think this is politically motivated research, because that couldn’t be further from the truth.”
| }
|