Exploiting Social Context for Anticipatory Analysis of Human Movement

Abstract

Our aim is to develop principled methods to transfer models of human movement using social context. The resulting techniques will form a fundamental contribution to the field of human terrain analysis, enabling diverse sources of data to be leveraged along with GEOINT and resulting in improvements to software tools used by analysts for the anticipatory analysis of human behavior. The philosophy behind our proposed approach is the following: it is more effective to represent an agent's social context with specific exemplars of people who share socio-cultural similarities than it is to create a parametric model over the entire population. Biased sampling techniques can allow the leveraging of large collections of data from groups of humans without necessitating the creation of an explicit model of interpersonal interaction effects. Existing human behavior models are best used to supplement data gaps. Our goal is to reduce the analysts' workload by identifying the relevant regions and timeframes in spatio-temporal data sets. This information can be used: (1) create intelligent data filters, (2) guide the future deployment of data collection capabilities, and (3) assess competing hypotheses. The information extracted using our techniques (augmented social networks, points of interests, reduced road networks) can be visualized and modified by the analyst to modify the search boundaries in an interactive fashion.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 01, 2012
Accession Number
ADA568418

Entities

People

  • Gita Sukthankar

Organizations

  • University of Central Florida

Tags

Communities of Interest

  • Autonomy
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Agent-Based Simulations
  • Bayesian Networks
  • Computational Complexity
  • Computational Science
  • Data Mining
  • Human Behavior
  • Information Processing
  • Information Science
  • Information Systems
  • Machine Learning
  • Monte Carlo Method
  • Probability
  • Social Media
  • Social Networks
  • Supervised Machine Learning
  • Surveys
  • Urban Areas

Readers

  • Computational Modeling and Simulation
  • Distributed Systems and Data Platform Development
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.

Technology Areas

  • AI & ML