24. Investigation of Key-Player Problem in Terrorist Networks Using Bayes Conditional Probability

In Social Network Analysis (SNA) it is quite conventional to employ graph theory concepts for example cut-points and cut-sets or measuring node centralities like betweenness, degree and closeness to highlight important actors in the network. However, it is also believed that most of these measures alone are inadequate for investigating terrorist/covert networks. In this paper we compute posterior probability using Bayes conditional probability theorem to compute each nodes positional probability to see how probable/likely that the node is a key player. Obviously, larger the probability value higher is the chances that the node is a key actor, the system after words using the computed probability can select an arbitrary number of nodes for elimination to make the network non-functional or severely destroying its capability. We know that that network fragmentation in its simplest form is a count of the number of components, more the number of counts larger is the fragmentation. The reason of calling it simple is because it does not include many aspects of the network for example structure etc. Borgatti has provided a similar measure which does includes the shape and internal structure of the components and high light the important actors in the network. Our computational procedure provides very similar results in highlighting the key player problem. We simulated our computational model through various random networks and we also applied it to the David Krackhardt’s kite network and to 9-11 hijackers network to bench mark our results. It shows that conceptually this framework of evaluation can be used to highlight the key players in terrorist/covert cells.