Disrupting Terrorist Networks Based on Link Prediction: A Case Study of the 9-11 Hijackers Network

Abstract

Relationships between terrorists are amorphous, invisible, distributed, and dispersed. The information on these networks is often incomplete and even erroneous. The key to disrupting the terrorists' network is to find the critical nodes whose removal will lead to network collapse, and however, most of the previous studies are based directly on the observed networks while neglecting an important fact that the observed networks may be incomplete. In this paper, we address the terrorist network disintegration problem based on link prediction. An effective method is proposed to find the critical nodes by the assistance of a link prediction algorithm. We make a case study of September 11th hijackers to illustrate our method. Five different disintegration strategies are applied to validate our method. The result shows that, with no more than 40% magnitude of missing information, by using the link prediction method to recover partial missing relationships information, our method can improve the network disintegration performance remarkably. Besides, we find that when the size of missing information is not too much, our method even outperforms than the results based on complete information. We refer to this phenomenon as the 'comic effect' of link prediction, which means that the original network has been reshaped through the addition of some links by link prediction. As a result, the reshaped network is like an exaggerated but characteristic comic of the original one, where the important parts are emphasized.