Steve Connor: Run the program, join the dots and find the terrorist

Social network analysis is a fancy way of "joining the dots" between different people who either know each other directly or who know someone who knows someone else in a web of contacts. It has been used as a management tool in business for years but has only recently been applied to terrorist cells.

The key moment for social network analysis probably came in the immediate aftermath of the 9/11 terrorist attacks on the United States in September 2001. It was then that civilian social network analysts began to investigate the 19 hijackers and their associates using publicly-available information released in the media.

Related articles

• 'Terrorist Facebook' – the new weapon against al-Qa'ida
• Taliban steps up election terror campaign
• Search the news archive for more stories

One management consultant, Valdis Krebs, scoured newspaper reports to build up a social network of the 19 terrorists that he could begin to analyse. Within weeks, his network began to build up a visual and mathematical picture of the links between the terrorists, with each name becoming a "node" in the net.

It soon became obvious to Mr Krebs that one of the key figures was Mohamed Atta, the Egyptian-born ringleader of the group who became a member of al-Qa'ida while studying in Hamburg. Atta's "node" scored highest on all three "centrality metrics" of the social network – measures known as "degrees", "closeness" and "betweenness".

"The network metric 'degrees' reveals Atta's activity in the network," explained Mr Krebs in a scientific paper on his analysis published in 2002. "'Closeness' measures his ability to access others in the network and monitor what is happening. 'Betweenness' shows his control over the flow in the network – he plays the role of a broker in the network. These metrics support his leader status."

The network also revealed the dense interplay of relationships between a subset of the hijackers who were later known to have formed the terrorist cell that Atta had helped to build up while in Hamburg. "The dense connections of the Hamburg cell are now obvious," Mr Krebs wrote in 2002.

The management consultant started building up the network from just two nodes: the two terrorist suspects, Nawaf Alhazmi and Khalid Almihdhar, who were photographed in early 2000 attending a meeting of known terrorists in Malaysia. After the meeting they returned to Los Angeles where they had set up residence in 1999.

Mr Krebs discovered from his social network analysis that all 19 hijackers involved in the 9/11 atrocity were within two steps in the network of these two original suspects uncovered in 2000. With hindsight, if that kind of information had been known before 9/11, intelligence agencies might have been able to prevent it.

The insight that this arcane study of human relationships brought to the fight against terrorism was not lost on the intelligence community. Mr Krebs and other social network analysts in academia and business were soon giving briefings to military chiefs in Washington on how it could be applied in the field.

Ian McCulloh, a US Army major at West Point Military Academy who has a doctorate in the subject, said that social network analysis has brought a mathematically-strict discipline to the art of traditional intelligence work.

"By representing a social network in a mathematical framework and by using computer algorithms we are able to find out the minimal number of nodes in a terrorist network needed to fragment the network and disrupt it," Dr McCulloh said.

"Largely, the 'intel' analysts know who these highly central people are, but what the network analyst can do is to highlight one or two important people who are not obvious," he said.

One of the next key developments is something called "metanetwork" analysis. This goes beyond analysing how individuals are related to one another in a network, to take in more complex associations that ask questions such as "when", "where" and "why", explained Professor Kathleen Carley at Carnegie Mellon University in Pittsburgh, Pennsylvania.

"Metanetwork analysis looks for links across many cases. It is qualitatively different. It lets you focus on the strategic thinking, not just the tactical thinking of the case. And I would say 'yes', it's working," Professor Carley said.

Dr McCulloh said that the US military is actively pursuing social network analysis and developments such as metanetworks. Two years ago, the US Army established a network science centre at a facility in Aberdeen, Maryland, and this year the US Army Research Laboratory has committed $162m (£100m) to collaborative research projects with academia and industry, according to the journal Science.

"Metanetworks is in my opinion the key important next step in what we want to do because what a social network analysis does is look at people and how they are related to others," Dr McCulloh said.

"But what you want to find out is who's important in this network."