Abstract
Since Sept. 11, 2001, there has been great interest in the military
and intelligence communities in using Social Network Analysis (SNA)
to support the disruption and destruction of global terrorist networks.
SNA results, however, tend to be descriptive and are limited due
to the lack of advantageous properties of the relationship measures
applied to the arcs in a social network. Further, SNA techniques
generally focus on a single network context while real relationships
are based in multiple contexts. This thesis develops a new proxy
measure of pair-wise potential influence between members of a network,
a Holistic Interpersonal Influence Measure (HIIM). The HIIM considers
the topology of the multiple formal and informal networks to which
group members belong as well as non-network characteristics such
as age and education level that may indicate potential influence.
The HIIM, once constructed results in a network of pair-wise potential
influence between group members. Further, the numeric properties
of the HIIM are appropriate for use in Operations Research Network
Flow models, which will enable analysts to provide prescriptive
analysis focused on specific actions and their outcomes. In addition
to an overall measure of influence, the HIIM methodology provides
important intermediate results such as the development of operational
group profiles The methodology is applied to open source data on
both Al Qaeda and the Jemaah Islamiah (JI) terrorist networks. Key
leaders are identified, and leadership profiles are developed. Further,
a parametric analysis is performed to compare influence based on
individual characteristics, network topology characteristics, and
mixtures of network and non-network characteristics
Links and resources
Tags