The Big K: Automatically Discovering the Most Informative Factors in Social Phenomena

Abstract

Humans are complex and many diverse and overlapping factors may affect any human decision. Social science research has generally been limited to small studies that attempt to isolate and test the effects of individual, human-hypothesized factors. In principle, the emergence of large-scale data about human behavior based on ubiquitous sensors and computing could allow us to automatically disentangle diverse factors behind human behavior in a principled, data-driven way. However, new tools are needed to realize this vision. The goal of this project is to automatically discover the most important factors that affect human behavior. Ideally, the number of factors, K, and the meaning of these factors could be discovered from data automatically, without human intervention. The factors we recover should be meaningful, reliable, and interpretable.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 19, 2019
Source ID
W911NF1610575

Entities

People

  • Greg Ver Steeg

Organizations

  • Army Contracting Command
  • Defense Advanced Research Projects Agency
  • University of Southern California

Tags

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Neural Network Machine Learning.
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.