The advent and spread of the global Internet and the subsequent dawn of the "Information Age" has brought scores of new economic and social opportunities. It has also exposed and exacerbated psychological, sociological, and organizational flaws in human societies. Recent history demonstrates that Internet-based communications and social structures present significant challenges to long-established social mechanisms and societal models. Left unaddressed, these issues will continue to destabilize existing social mechanisms.

Addressing the challenges of the Information Age has proven difficult for policymakers and leaders for many reasons. Information is an inherently non-geographic quantity and is intrinsically produced by, linked to, and leveraged in, all human activities. In addition, information is a causal driver of human behavior. The ability of policymakers to understand and affect the information available to constituents, in competition with other information providers, determines how effectively they can govern. The lack of domain expertise makes leaders dependent on models to understand and direct information activities, but available modeling methods are insufficient for the rapidly evolving information environment.

This thesis presents a novel model of the information environment offering high- level leaders, such as state policy makers or military leaders, a tractable, intuitive, and helpful representation of the information environment. The model is designed to capture and leverage the interconnectedness of different agents, topics, and platforms, to help forecast the effects of proposed changes and decisions.

The thesis incorporates and extends the BEND influence framework into the BENDRS framework, providing a lexicon to relieve policy makers of need for a detailed understanding of the specific means and methods of information operations. By abstracting complex edge-level concepts in cyber, propaganda, and covert influence, the lexicon helps leaders recognize, understand, and direct changes in the environment, without requiring them to prescribe the means and details of effecting that change. Thus, it supports a state-type hierarchical organization composed of many varied fields of expertise and charged with a highly heterogeneous purview and a broad and evolving set of goals.

Finally, the thesis demonstrates methods to construct such a model solely from observed network traffic, making it immediately operationally accessible and removing any requirement for an omniscient network architect. The network is constructed from relationships and values observed within communications traffic, and is iteratively and additively adjusted as additional traffic is revealed; assuming all traffic is available and intelligible, the derived network, in the limit, converges to the actual network environment. More importantly, the derived network in any state is immediately operationally useful as a quantitatively analyzable snapshot of the current environment.

225 pages

Thesis Committee:
Kathleen M. Carley (Chair)
L. Richard Carley
Geoff Dobson
Nathan VanHoudnos (Software Engineering Institute)

Nicolas Christin, Head, Software and Societal Systems Department
Martial Hebert, Dean, School of Computer Science

Return to: SCS Technical Report Collection
School of Computer Science

This page maintained by reports@cs.cmu.edu