Quantitative Modeling of Rumors

From MilcordWiki

Overview

Rumor is unconfirmed information circulating among persons endeavoring to make sense of a situation that is ambiguous or potentially threatening. Like news, rumor is of current or topical interest and is generally considered important or significant; unlike news, it is never verified. Participants may or may not be aware of the rumor’s unstable foundation—they may regard it as fact. Nevertheless, doubtful or deficient supporting evidence is rumor’s central defining feature. The United States, through combat deployments, peacekeeping operations and humanitarian responses, comes face to face with the phenomena of rumor in cultural contexts that are challenging to comprehend.

Need

Rumor can impede and hinder operations, while at the same time providing insight into the drivers of fear and hostility for specific target groups. Understanding the rumor environment can help provide situational awareness to evolving socio-cultural dynamics relevant for operations. Therefore tools for detecting and tracking rumors, and the development of frameworks for assisting operators to better understand the meaning behind rumors, pathways of rumor propagation, and concepts of operation to counter rumors are important for supporting full spectrum operations.

Approach

Rumor modeling can answer questions related to influence operations such as: How does a specific rumor spatiotemporally travel over the area of operations? Will this rumor repeat itself? Can rumors predict sentiment survey results? Can rumors predict incidents? What kinds of rumors have the most influence on what type of events? What is the time lag between rumors and corresponding realization of actual events?

As an example, consider the belief network model developed using data from Baghdad Mosquito with the rumor coding in [Kelley 2004] . The figure on the left shows the resulting belief network model for the targeted rumor. As expected, ‘subject’ influences ‘rumor’ and ‘target’ while ‘rumor’ influences target. The model shows that ‘subject’ influence on ‘target’ is an order of magnitude higher than the ‘rumor’ influence’ on ‘target’. We use this model framework to study the influences of various rumor campaigns. For instance, when the subject is ‘quality of life’, the only targets are the US forces and Iraqi government while the ‘fear’ overtakes ‘hostility’ in beliefs.

Benefits

Understanding rumors can help operators more effectively develop tailored communication campaigns across strategic, operational, and tactical levels that can assist in reducing anxieties before rumors develop, discover and alleviate the underlying causes related to rumor development, and minimize damage rumors cause. Developing a targeted information campaign can promote counter-extremism narratives as well.

Applications

• Military: Influence and military information support operations
• Civilian: Counter-extremism campaigns
• Commercial: Brand sentiment tracking

References

• Nel, F.; Lesot, M-J.; Capet, P.; Delavallade, T. (2010) Rumour Detection in Information Warfare: Understanding Publishing Behaviours as a Prerequisite, NATO RTO-MP-IST-091 Information Assurance and Cyber Defence.
• Guttieri, K., and Caglayan, A. (2009) Rumour and Perception amidst Social Conflict, European Consortium for Political Research, 5th General Conference, Potsdam, Germany, September 10-12, 2009.
• DiFonzo, N., and Bordia, P. (2007). Rumor Psychology: Social & Organizational Approaches, Washington, DC: American Psychological Association.
• Kelley, S. R. (2004) Rumors in Iraq: A Guide to Winning Hearts and Minds, Masters Thesis. Naval Postgraduate School, Monterey, CA.