Tuesday, November 30, 2010

Social Network Analysts Fight Terrorism

Since the 9/11 terrorist attacks, the United States and their alliance partners have been trying to destroy the responsible terrorist network named Al-Qaeda. After initial successes with the use of military power, the U.S. came to the conclusion that this means alone is not sufficient. Al-Qaeda is not a militarily structured unit and avoids battles that it cannot win. Additionally, to take out just the obvious elements – the fighting members of the network - will not destroy its underlying structure, which consists of parts such as leadership, training and all kinds of resources, for example material, weapons, communication equipment etc.

Social network analysis (SNA) can be used to collect crucial information from such a network. After 9/11, Valdis Krebs created a map to show the ties between the involved hijackers. Like Krebs, the Spanish sociologist Jose A. Rodriguez used public sources to draw a map visualizing the connections of the March 11th network, after the bombings in Madrid in 2004. Further analyses regarding Al-Qaeda were made by Marc Sageman. He used 172 biographies of known terrorists from different networks to show the ties among them. He discovered four geographic clusters of the network: around the Afghan-Pakistan border, in some Arab states, in the Maghreb in North Africa and in Indonesia and Malaysia. Many models of the network were created by academics but they were of limited use since the availability of trustworthy and complete data was not always sufficiently given.
The U.S. government has much more and better information than those entities using only publicly available data. Organizations such as the Central Intelligence Agency, the National Security Agency and the Department for Homeland Security use social network analyses.

Assuming that we have sufficient and correct data, how can SNA contribute to destroy a terrorist network better than other means? Nodes of the network with high centrality can be identiefied. They are potential points of failure and should be removed. It is favorable to eliminate them at the same time so that there is no time to replace nodes. Furthermore, individuals whose removal would disconnect groups can be identified. It is not just about to remove as many as you can but about to remove the important individuals. With that, a quicker implosion of the network is much more likely. And it is also a matter of time and resources one wants to spend on observation of the individual nodes of a network. Fellman and Wright call it to get the “most bang for the buck” - to cause the biggest possible damage with the given resources. Additionally, in groups were each member has strong ties to many other members, removing their leader would be of little effect. In contrast to that, members with a high connectivity and special skills are the weak points of a network.

After 9/11 SNA helped to prosecute terrorist. Unfortunately it could not prevent the attacks because these networks were hardly visible, since there were no ties between the hijackers and outsiders. But due to the strong ties within this network, the connections between them could be quickly identified after the attack.

Fellman P., Wright R., Modeling Terrorist Networks – Complex Systems at the Mid-Range, http://www.psych.lse.ac.uk/complexity/Conference/FellmanWright.pdf, viewed: 30 November 2010

Krebs V., Uncloaking Terrorist Networks, FM
Volume 7 Number 4 - 1 April 2002, http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/941/863, viewed: 30 November 2010

Ressler, S., Social Network Analysis as an Approach to Combat Terrorism: Past, Present, and Future Research, Homeland Security Affaris, 2006, http://www.hsaj.org, viewed: 30 November 2010

3 comments:

Christopher Tunnard said...

Nice post, Captain Krause. And as you can imagine, the analysis has gone way beyond Krebs's original 9/11 network map. But that map was the start, indeed. Many authors (including me) have focused on how SNs can help form resistance networks, but, as Fellman and Wright point out, SNA can help governments find the fragile points of networks even more easily. For example, I have colleagues working on the Taliban networks in Afghanistan and Pakistan right now, but not just to find military vulnerable points but political ones as well.

Michael Mars said...

Interestingly, the SNA analysis used in counterterrorism has shown that the key actors in a terrorist group, which have usually high level of connectedness, are seemingly low-level people, such as drivers, who keep addresses and phone numbers memorized. In order to find these, analysts trace short phone calls placed to the same number just before and after an attack. For example, the mapping of the social network of Saddam Hussein's chauffeurs has played a key role in his capture in 2003 on a farm near his hometown of Tikrit. This according to the CEO of i2, a British firm which developed the software used in the manhunt.

Further, network analysis is now used for forecasting models. For instance, one such model analyses information about politics, business and society in Lebanon to predict rocket attacks by Hezbollah on Israel. It has found that attacks increase when more money from Islamic charities flows into Lebanon, and decrease during election years.


Reference:
The Economist - Technology Quarterly, Sep 2010

Christopher Tunnard said...

Yes, I know both these cases. In fact, the flow of funds has proven to be the best indicator of terrorist activism, since most of the operational communications have gone over to SMS on throwaway, or burn, mobiles. Almost impossible to trace. There are indications that mobile banking is now becoming popular as well, for the same reason.