Human Networks: From Algorithmic Models to Global Behavior
Abstract
The major goal of this project is to determine to what extent complex interactions between human agents can be modeled and predicted using the mathematical tools offered by the engineering discipline. We want to explain and predict large-scale outcomes observable in a population of intelligent agents using simple models of local interactions. Despite their simple rules of interaction, these agents must have the descriptive power of the real population, and explain large-scale observations. The model should also be prone to rigorous mathematical analysis, allowing us to make predictions. When compared with real data, the model should be able to explain how global behaviors can be triggered by events and conditions that would not be apparent in controlled, small-scale experiments. On-going work includes models of segregation and population dynamics, spreading of viral epidemics, and cooperation algorithms. Currently, our approach uses mathematical models and analysis of data from massive online social networks. Our preliminary results in this first phase of the project marked important steps towards the understanding of the collective dynamics of complex human systems and can provide relevant guidelines in a variety of strategic decision-making scenarios. These include understanding of adversarial networks triggered by segregation models, revolutionary triggers, and sustainable societal norms.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Sep 11, 2018
- Source ID
- W911NF1510253
Entities
People
- Massimo Franceschetti
Organizations
- Army Contracting Command
- United States Army
- University of California, San Diego