Understanding Social Influence without Markov Assumptions - Research Area 10: Network Science

Abstract

The diffusion of ideas, information, or behavior through a social network has been modeled in many different ways. The majority of these models assume that an individual in the network receives a social contagion immediately after one or more of his or her neighbors have received the condition. However, there are two shortcomings in these models. First, they implicitly make the Markov assumption: an individual receives a contagion based on the state of his neighbors in only the previous time-step. Second, they assume that other precursor contagions do not change the speed and/or likelihood of adoption. In this paper, we propose to model social influence without Markov assumptions while allowing for precursor contagions using a two-pronged approach; we will: (1) develop a temporal-logic based framework for social network diffusion that allows for non-Markov relationships as well as precursors and (2) extract diffusion models under such a framework from human-subject testing using Sandia National LabsÕ ÒControlled, Large, Online Social ExperimentationÓ (CLOSE) platform. With our approach, we anticipate not only achieving a much fuller understanding of the spread of social influence but also developing the computational tools to learn and reason about such processes for a real-world application.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 19, 2019
Source ID
W911NF1510282

Entities

People

  • Paulo Shakarian

Organizations

  • Arizona State University
  • Army Contracting Command
  • United States Army

Tags

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Mathematical Modeling and Probability Theory.
  • Systems Analysis and Design