Culture Sensitive Predictive Modeling of Societal Instabilities
Abstract
Recent events in volatile geo-political regions underscore the need for quantitative models and theories to enable predictions of emergent properties of social systems, such as social dynamics, influencing, socio-political instabilities, and tipping points. We propose to investigate how to capture sweeping changes in cultural, religious, and political beliefs and affiliations with featurerich individual-based models in which individuals are shaped by the culture of the modeled society. We will focus on features that impact human behavior most profoundly, such as the influenceability of individuals, their attitudes to novelty and change, the scope and range of their connections. Such features may affect dynamics of opinion spread and the social networks that link them together and are measurable from data collected from the existing dynamic social networks. The states of the nodes (individuals) in social networks are governed by pairwise- (or small-group) influencing and external (global) media effects. Yet, in a socially interacting population, the evolution of links is driven by the theory of structural balance that is one of the key driving mechanisms of social dynamics. This theory has been studied extensively in the context of social networks, but not when it competes simultaneously with other social drivers. In our work, we will consider models for pairwise- (or small-group) influencing (based on our past work on Naming Game and linear threshold models, both with their feature rich variants) and models in which the driving force comes from structural balance. To better understand the competing effects of fundamental social drivers, one needs data-driven models with predictive power where individuals’ complexity ranges from simple to feature-rich interacting agents, giving rise to traceable changes in observables at the global and societal levels. The outcome of these competing processes depends crucially on the connectivity and interactions between individuals (including online social networks today) and on cultural aspects such as strength of family, tribal, political and national ties, attitudes to change and novelty, range and scope of social ties, as well as on the availability and intensity or the suppression of community-wide or global information (e.g., news and media) for individuals/citizens. The proposed research shall develop fundamental theoretical quantitative modeling approaches to describe the complex interrelation of cultural and political affiliations and institutions as they affect societal stability. The results will not only advance fundamental knowledge, from a network science viewpoint, but will also create methodologies and techniques that could be applied to address strategic and urgent national priority aspects of this research. Using a novel combination of stochastic individual- and agent-based models for opinion formation and influencing and empirical social graphs, research will develop models with predictive power for large-scale social networks. In the model tuning stage, we will seek data with measures of population attitudes and perception on issues, people, and events from social media collected from different sources, such as sentiment analysis from blogs, tweets, and Facebook. If available, we will use sentiment propagation across the internet and identification of key sites and people in shaping opinion dynamics to tune parameters of our models to specific societies. As the result, those models shall provide an array of strategies for "what if" scenarios that will enable analysts to answer such questions as where and how to defend networks with neutral or tolerant "ideologies" against militant infiltration, and conversely, where and how to attack (by identifying tipping points) and disintegrate adversarial communities exhibiting hostile, extremist and/or militant ideologies.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 12, 2016
- Source ID
- N000141512640
Entities
People
- Boleslaw Szymanski
Organizations
- Office of Naval Research
- Rensselaer Polytechnic Institute
- United States Navy