Data analysts and “information cascade” experts are using their skills to frustrate the spread of terrorist ideology from groups like ISIS, who grow increasingly adept at online recruitment.
It’s difficult to equate the image of a group like ISIS, which on the one hand wishes to restore a medieval-style extremist Islam upon the 21st century, while on the other is a tech-savvy movement capable of doing a significant proportion of its recruitment online. ISIS has proven itself to be an efficient user of social media, particularly for the enlistment of those alienated individuals around the world searching for a sense of belonging.
So much so that a leaked document from the U.S. State Department in June admitted as much. The New York Times reported from an internal government memo describing how ISIS had “trumped” attempts by several of the world’s most technologically advanced nations to thwart its progress online.
Now research from the CyberSocio Intelligent-Systems Laboratory — conveniently shortened to CySIS — at Arizona State University (ASU) brings together both online and offline expertise from social science and data analysis to help stall the spread of online extremist ideology.
“We are developing better tools to detect extremist networks promoting violence and block their online content,” explains Dr. Hasan Davulcu from the School of Computing, Informatics, and Decision Systems Engineering at ASU.
“Online recruitment by groups like ISIS tends to be based on appeals to alienated individuals by promoting utopias as pull-factors and anti-western sentiments as push factors,” Davulcu stated. “Once we identify the core pull factors, we can block their online content promoting violence, amplify voices that point to inconsistencies between their ideology and the reality, and bring to view alternative voices who present a vision of a more peaceful, tolerant and progressive world.”
The best way to fight a network is with another network. Dr. Hasan Davulcu
The analysis of “information cascades” is a key online research tool for CySIS. “The spread of messages, memes or news items that go viral can have a structural diversity which is telling,” explains professor Paulo Shakarian from CySIS. “Firstly, a message that disperses into a variety of online communities is more likely to go viral. For example, if you were to receive the same tweet from three work colleagues, that’s only really one source and so is unlikely to spread much farther. If, on the other hand, you were to receive a tweet from a family member, a work colleague and an old college friend, that’s potentially more significant. We have developed metrics to assess the significance of how a message or micro blog spreads online.”
Shakarian will work closely on this five-year project with Davulcu, who has developed a tool called Looking Glass software, which is designed specifically to generate, source and deliver the most extensive and digestible info on various threat data online.
“We have developed new computational representations, algorithms, and tools allowing analysts to achieve unprecedented precision and coverage of complex socio-cultural phenomena,” says Davulcu.
The graphical tools provided by LookingGlass are complemented by expertise from people on the ground in various regions across the world.
“Initially, we work with area experts to identify on-the-ground groups,” says Davulcu. “Next, we find their online media and followers, and we track the correspondence between what goes viral online (e.g., the issues, events, perspectives), and explore the correspondences between virtual observations and on-the-ground effects.”
This is an ongoing research project at ASU, and new funding has just been secured. The previous five-year study included various social media and relied upon culturally contextual input from experts working across three continents in nine countries. The next five years will focus on Twitter use, specifically in Iraq, Syria, Nigeria, Indonesia, Malaysia and the U.K.
“We chose the UK because of the large minority populations there but also to give more cultural context,” says Shakarian. “People in different countries use Twitter in different ways. The way messages are sent is culturally salient in terms of our analysis. That’s why the field research is so important. It helps us go beyond the information cascade results and provide a more comprehensive understanding of the data.”
Shakarian is considered the world’s leading “information cascade” analyst. A graduate in Computer Science from West Point Military Academy, his expertise and the research generated from CySIS more generally will help inform U.S. military intelligence, as well as other allied forces fighting the spread of terrorist ideology online.
They may end up finding allies in more unexpected places, too. Hacktivist group Anonymous has recently increased its own efforts to curb the spread of the Islamic State’s message online.
Over the summer, Anonymous published a list of Twitter accounts being used by the group, while flooding others with unrelated images in an effort to affect search engine results. Other sites have been suspended or shut down as a result of the hactivist’s campaign.
“Anonymous has been attacking and hacking the websites of ISIS for a long time now,” says Davulcu. “They already understand that the best way to fight a network is with another network.”