September 16, 2011

Researchers Skeptical DOD Can Use Social Media to Predict Future Conflict

Can Twitter predict insurgent attacks?

The Pentagon wants to know, but one group of university researchers already has an inkling of the answer: Doubtful.

From a military standpoint, there is a big question about how the Defense Department or intelligence agencies can use social media as a tool. Can the Pentagon track developments of an ongoing conflict or political upheaval and influence them, or maybe protect Americans at war from attack, by analyzing real-time data revealed within tweets or Facebook friend connections and blogosphere locations?

U.S. spy agencies really, really want to track social media. Intelligence Advanced Research Projects Activity recently put out a solicitation to figure out how to use “Open Source Indicators” the way people predict hurricanes. 

But hold those horses.

“At least in this group here, there’s a fair amount of skepticism about the ability to make pointed predictions about insurgent attacks or insipient protest or riots by tracking social media,” said Marc Lynch, associate professor of political science and international affairs at George Washington University.  “Where there’s more success is in shaping attitudes and narratives.” 

At the U.S. Institute of Peace’s “Sifting Fact from Fiction: The Role of Social Media in Conflict” event on Friday, Lynch and others who have studied this as part of their “Blogs and Bullets” program described in-depth research about Twitter, blogs and other media during the first few months of the Arab Spring. 

“The new media are a megaphone,” said John Sides, GWU political science profressor and contributor to The Monkey Cage.

New media, he said, serve to push out event news, feelings and attitudes to foreign audiences, helping organize the protest movement (January 25 at Cairo’s Tahrir Square) but then disappearing when the event seemingly ended (Mubarak’s resignation).

The group studied every tweet about Libya and Egypt that contained a web link shorted by bitly and identified the time and location of every person who clicked on those links. Roughly 75 percent of the 1 million clicks on Libya-related tweets and 89 percent of the 3 million clicks for Egypt-related Tweets came from outside of the Arab world. 

Among the most popular links were to live-streaming video from the al-Jazeera broadcast television feed, Sides said, but then again, so was The Guardian’s famous quiz on whether quotes were from Mohamar Gadhafi or Charlie Sheen.

Foreigners also had a somewhat adverse affect by flooding Twitter with hashtags about the events, which tended to drown out the actual Twitter users in the Arab World.

They also found the foreign audience has “a very short attention span.” When Mubarak resigned, foreign traffic dropped significantly.

After the revolution (or, more accurately, post-regime collapse), Twitter has not held the protest movement together or aided in building new institutions required to replace the regime (think dissidents in Poland). Instead, groups like Muslim Brotherhood that already was organized with institutional infrastructure and had very little Twitter use during the protests, have taken advantage of the post-conflict power void.

So, what can soldiers and war planners back in Washington do with the social media lessons being learned from the Arab Spring?

According to Lynch, as there is broader recognition and appreciation of the importance of social media for creating attitudes, narratives and identities, social media becomes “an arena within which to compete, to try and undermine extremist narratives or to try and get the message out.”

“I think it’s inevitable that all governments, militaries, and intelligence agencies are going to get into the game,” he said. “They’re going to see that it matters and they’ll want to be there.”

That was on rare display this week as NATO and the Taliban actually had a tweet warover the Kabul attack.

Monitoring social media is one thing. Using it to influence political events is another, and the panel offered a clear warning to the U.S. government: Don’t.

In 2009, in the middle of the Iranian protests, Twitter lit up with solidarity messages from around the world, meant to aid the protest moment’s organization and circumvent Iranian censorship. The U.S. urged Twitter to not go through with a scheduled weekend maintenance outage. That move was a mistake, according to Henry Farrell, associate professor of political science and international affairs at George Washington University who studies politics and the Internet.

Few protestors in Iran were actually using Twitter, compared with the foreign audience for it. Instead, the U.S. move drew unnecessary attention to Twitter and allowed Iran to portray it as a tool of the U.S. government.

“Direct efforts by the U.S. to leverage this are likely to result, at best, in somewhat problematic situations,” he said, “at its worst in disaster.”