Opponents of the Islamic State in Iraq and Syria are six times greater in number on Twitter than ISIS supporters, but those sympathetic to the group are more active on the social media platform, according to a new RAND Corporation study.
Analyzing more than 23 million tweets posted in Arabic over a 10-month period, researchers found that supporters of ISIS produce 50 percent more tweets than opponents on a typical day, although there is evidence that ISIS opponents are increasing their activity.
Researchers say that U.S. officials should do more to support opponents of ISIS on Twitter, possibly offering social media trainings and other engagements to enhance the effectiveness and reach of their messaging.
"Organizations such as the U.S. military and the State Department looking to counter-message ISIS on Twitter should tailor messages and target them to specific communities," said Elizabeth Bodine-Baron, the study's lead author and an engineer at RAND, a nonprofit research organization. "The ISIS Twitter universe is highly fragmented and consists of several different communities with different concerns, so messages need to be aimed at specific audiences, rather than trying to craft a one-size-fits-all message."
Like no terrorist organization before, ISIS has used Twitter and other social media channels to broadcast its message, inspire followers and recruit new fighters. Though less heralded, ISIS opponents also have taken to Twitter to denounce the ISIS message.
RAND researchers used a variety of methods to analyze 23 million tweets posted by 771,327 users from July 2014 to April 2015. The findings allowed researchers to identify more than 20,000 distinct user communities and group those into four major meta-communities that characterize the conversation about ISIS on Twitter.
Those four meta-communities include: Shia (they generally link ISIS to Saudi Arabia but express support toward the international coalition fighting ISIS), Sunni (the most fractured group, with several sub-communities focused on country-specific issues), Syrian mujahedeen (opposed to the leadership of Bashar al-Assad in Syria with mixed attitudes toward ISIS) and ISIS supporters (this group frequently invokes threats against Islam, characterizes its enemies as "other" and employs social media strategies such as actively encouraging sympathizers to "spread" messages to expand their reach).
Though fragmented, the patterns of connection between the communities opposed to ISIS suggest inroads for influence that the U.S. government's social media strategy should explore in order to weaken the ISIS Twitter propaganda and online recruitment, according to researchers.
The study showed that near the end of its reporting period (spring 2015), the number of ISIS supporters active on Twitter decreased while the number of opponents increased. This change coincided with Twitter's campaign to suspend the accounts of ISIS supporters. Researchers also found that ISIS atrocities such as the burning of the body of a Jordanian pilot sparked a huge upsurge in anti-ISIS tweets.
"Twitter should continue its campaign of account suspensions that harasses ISIS Twitter users, forcing them to lose valuable time reacquiring followers, and ultimately may push some to use channels that are far less public than Twitter," said Todd Helmus, an author of the report and a RAND senior behavioral scientist.
The study used different network analysis tools and algorithms to identify and characterize the conversation on Twitter about ISIS. For example, researchers found that they could separate supporters from opponents using a simple method: ISIS supporters typically refer to the organization in Arabic as the "Islamic State," whereas opponents typically use the disrespectful Arabic acronym "Daesh." They say the method can continue to be used to gauge the worldwide activity of supporters and opponents of ISIS.
The report, "Examining ISIS Support and Opposition Networks on Twitter," can be found at http://www.rand.org. Other authors of the report are Madeline Magnuson and Zev Winkelman.
Funding for the study was provided, in part, by donors and by the independent research and development provisions of RAND's contracts for the operation of its U.S. Department of Defense federally funded research and development centers.
The research was conducted within the RAND National Security Research Division (NSRD). NSRD conducts research and analysis on defense and national security topics for the U.S. and allied defense, foreign policy, homeland security, and intelligence communities and foundations and other nongovernmental organizations that support defense and national security analysis.