How a Computer Model Could Help Fight Terrorism
Modeling the flow of information in terrorist networks may help shut them down.
When Justin Bieber tweets, 39,361,876 people (and counting) immediately jump to attention. But when one of those nearly 40 million people tweet, does the Beebs see it? Does he react at all?
Communication among terrorist cells works much in the same way as Justin Bieber's Twitter account, according to mathematicians from Ryerson University in Toronto who have built a mathematical model of the way information spreads through these hierarchical networks. Their approach may give counterterrorism agents insight into terrorism hierarchies and allow them to predict terrorist attacks and sabotage networks before the attack plans can be carried out.
Terrorist networks are often arranged hierarchically, meaning information flows in one direction: top down, from one leader to many followers. This model is called a "directed network without cycles," or a "directed acyclic graph." [See also: Google Search Algorithm Models Cancer Spread]
The limited and one-way nature of the contact among parties in these types of networks helps preserve anonymity, and makes it easy for terrorist leaders to blast messages out to a large audience. What these "directed networks without cycles" lack in teamwork they make up for in sheer numbers — leaders can assume that within their huge networks, at least one person will act on their commands.
But these hierarchical social networks have serious flaws, which counterterrorism agencies could use to detect and even sabotage terrorist networks before they succeed in carrying out an attack.
In a paper describing their mathematical model, the researchers compare the way information flows top-down to the way lava flows down a volcano's side. There's no practical way to contain the lava from every possible point — but you can minimize the damage by blocking the lava's flow at a handful of strategic points.
The key advantage of this model is its flexibility: It's able to account for the slow spread of information over time, and also gives counterterrorism agents — the ones blocking the lava's flow — the ability to respond dynamically as new pathways present themselves.
Granted, the model does operate on a number of assumptions, including that the hierarchical social structure is consistent throughout the terrorist network.
And there will always be rogue actors who act in unpredictable ways.
"The Boston bombers are a good example of how little we know about such terrorist networks," acknowledged Anthony Bonato, a mathematics professor at Ryerson and a co-author of the paper. "Did the Tsarneav brothers act alone, or as part of a more extensive network? Further, the structure and organization of these networks are not well understood."