Our Top Ten Blog Posts by Readership in 2011
Originally published on September 6, 2011
Most of those who have been riveted to the breaking news in North Africa and the Middle East during the so-called “Arab Spring” and the recent grimmer months this summer have been focused on predicting the actions of the various heads of state—of Hosni Mubarak, Muammar Gaddafi, Bashar Assad. But many academics have been trying to figure out who have been the prime movers of the grassroots unrest sweeping the region.
Specifically, researchers have been looking into what roles social media have played in mobilizing opinion as well as galvanizing the revolutions in the streets. Academics have wanted to identify who has been talking to whom, as well as learn who have been the most influential voices. Whose messages are the ones passed on?
Among researchers, Twitter has been the medium of choice to investigate. Twitter launched on March 21, 2006, and three years, 2 months and 1 day later it sent its one-billionth tweet. Now Twitter handles a billion tweets every week. The exponential growth of the service as well as the relative ease with which researchers can collect tweets and organize them by “followers,” “retweets” and “mentions,” has made it a data trove for those who seek to calibrate influence and map social and political networks. (If you are new to Twitter, you can find an overview of how it works here).
Interestingly, what researchers are finding, however, is that the most popular individuals and groups —those who have the most followers—are not necessarily the most influential. In a paper published last May, for example, researchers from the Max Planck Institute for Software Systems discovered that what matters more is who is “retweeted,” because “retweets…could propagate multiple hops away from the source before they are retweeted throughout the network.” In a paper titled “The Million Follower Fallacy,” the researchers also observed a reinforcing mechanism at work: “retweeting in a social network can serve as a powerful tool to reinforce a message—for instance, the probability of adopting an innovation increases when not one but a group of users repeat the same message.” And finally, they noted that those who are retweeted often gain their influence “through concerted effort such as limiting tweets to a single topic.”
Building on such insights, Jorge Faytong Real and Nishant Patel, two graduate students at the University of Maryland working on a project in Prof. Ben Shneiderman course on Information Visualization, took a look at three slices of the Twitter universe to determine who have been “influentials” during the Middle East unrest. Faytong and Patel looked at the Twitter network of an extraordinarily-well linked U.S-based Syrian and Middle East activist who tweets in both English and Arabic: Ammar Abdulhamid—known to Twitter users as @tharwacolamus.
As they mapped the network of the several thousand followers of @tharwacolamus, Faytong and Patel noted that the most active 300 followers (with an average of 5,300 tweets each), in turn had a non-unique aggregated total of over 17 million followers—a size that made this whole “an excellent representation of the Middle East activist network.” (Note that Faytong and Patel conducted their network analysis and generated their data visualizations with NodeXL, a free, open-source tool).
Then Faytong and Patel overlay the Twitter networks of several mainstream U.S. media outlets, to see how well they were connected to Middle Eastern sources. To their great surprise, many U.S. mainstream media, including CNN, ABC News and US News and World Reports (see the lines in red),
“have very few primary sources with respect to Middle Eastern activists. On Twitter, they mostly rely on secondary sources, which are often based outside of the Middle East, implying a mutual lack of interest. Data shows the Middle Eastern activists are indifferent towards these media outlets. One possible cause is that these activists cannot depend on U.S. media outlets for timely and richly detailed reports. Their traditional news retrieval architecture results in delays in reporting developments from that region and watered-down analyses. This in turn keeps them well-buffered from the activists that are the most-connected.”
As Faytong and Patel cogently observed, American mainstream media outlets’ “interest in Twitter as a method of communication and information sharing is minimal. For example, @cnn has 1,886,456 followers, but only follows 538 users, meaning that it does not factor this social networking tool into its news retrieval infrastructure.”
Finally, Faytong and Patel explored the networks of alternative news outlets to see if those had strong connections in the Middle East. To their surprise they discovered that the Huffington Post was one of the best-linked channels on the Middle East protests [mapped with the top matrix of lines in red]. As they explained,
“The Huffington Post is an online blog that aggregates content from a wide variety of sources on a single site. It differs from traditional media outlets by relying largely on third-party contributions. This has created a niche for voices who do not receive proper attention nor have access to traditional media outlets, in this case Middle East activists. Its Web 2.0 architecture has created and continually strengthens a symbiotic relationship with its worldwide followers who can also participate as contributors.”
Why does all this matter?
The research matters because, as the graphics dramatically show, if news outlets—and policy makers— really want to get the pulse of a community, they need to look for and listen to those individuals and groups whose are connected in dynamic ways. They need to look for and listen to those individuals and groups whose messages are being passed on—or in the case of Twitter, those whose messages are being retweeted.
The common way to identify who matters in a community is to look for who’s popular, who is followed, or even, who has the most votes. The new research on Twitter suggests that media and policy makers need also to look for those who have a track record of saying things that others find valuable to pass on.
In fact if policy makers and media are NOT listening to those kinds of “influentials,” then their projections of the “Gemeinschaft” of a community may be appreciably skewed, and their policy recommendations may ultimately fall short of what is needed or wanted.
 This week, for instance, the New York Times wrote about a “provocative thesis of a new paper,” titled “Media Disruption Exacerbates Revolutionary Unrest,” by Yale political science graduate student Navid Hassanpour. Hassanpour argued in his paper that President Hosni Mubarak’s decision to turn off the Internet and shut down the cellphone service on January 28, at the height of the uprisings, was “not so smart, but not for the reasons you might think.” The reason why that decision was a mistake, argued Hassanpour, was that “Full connectivity in a social network sometimes can hinder collective action.” Or, as the Times summarized it: “all the Twitter posting, texting and Facebook wall-posting is great for organizing and spreading a message of protest, but it can also spread a message of caution, delay, confusion or, I don’t have time for all this politics, did you see what Lady Gaga is wearing?”
 Meeyoung Cha, Hamed Haddadi, Fabricio Benevenuto and Krishna P. Gummadi, “Measuring User Influence in Twitter: The Million Follower Fallacy,” presented at the Association for the Advancement of Artificial Intelligence’s International Conference on Weblogs and Social Media, May 2011
 Other important investigations of roughly the same public space date back several years. In 2009, for example, Bruce Etling, John Kelly, Robert Faris, and John Palfrey at Harvard’s Berkman Center published a paper titled: “Mapping the Arabic Blogosphere: Politics, Culture, and Dissent” that looked at approximately 35,000 Arabic-language blogs in order to assess the networked public sphere in the Arabic-speaking world. (That paper was adapted for publication in Dec. 2010 for the journal New Media and Society: http://cyber.law.harvard.edu/publications/2009/Mapping_the_Arabic_Blogosphere and http://nms.sagepub.com/content/12/8/1225.short.) The Harvard researchers found that that conversations occurred in national clusters, most prevalently in Egypt, Kuwait, Syria, and the Levant, as well as what they called “an ‘English Bridge’ group.” They also determined that the bloggers paid greatest attention to domestic political issues; although, as they noted “concern for Palestine is the one issue that unites the entire network.” Etling and his colleagues also observed that bloggers most linked to the most familiar Web 20 sites, such as YouTube and Wikipedia, and then were most likely to link to pan-Arab mainstream media sources, such as Al Jazeera.
 As of August 29, 2011, @cnn had 2,502,074 followers, but only followed 535. Note too that CNN is hardly unique in the disparity between their number of followers and the number of Twitter users they follow. @AJEnglish (Al Jazeera English) has 529,280 followers, but only follows 164, @nytimes has 3,665,290 followers, but only follows only 515, and @bbcnews has 312,901 followers, but only follows only 50. It is more typical that individual journalists for those news outlets would have a greater parity between the number of followers they have and the number of “sources” that they follow.
 They further noted: “@huffingtonpost has nearly a million followers and is the 355th most popular Twitter account in the world (far higher if you do not count entertainment celebrities).”
Photo Credit: Flickr user Brajeshwar