CARA: Cognitive Architecture for Reasoning About Adversaries

Abstract

We developed the CARA (Cultural Adversarial Reasoning Architecture) that provides a software platform using which analysts can reason about cultural groups, tribes, and terror groups. We developed the T-REX system to automatically extract information about violent events from news sources worldwide, as well as Stochastic Opponent Modeling Agents that allow us to reason about adversarial behavior. We developed two algorithms to make forecasts about adversarial behavior. We developed game-theoretical approaches to reason about group behavior. We tested out our approaches with real world data related to Hamas and Hezbollah and made forecasts that were true even though the terror group in question knew about the forecasts.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 20, 2012
Accession Number
ADA563483

Entities

People

  • Dana S. Nau
  • V. S. Subrahmanian

Organizations

  • University of Maryland

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Algorithms
  • Artificial Intelligence
  • Computational Science
  • Computer Science
  • Ethnic Groups
  • Game Theory
  • Human Behavior
  • Information Science
  • Intelligent Systems
  • Multiagent Systems
  • National Security
  • Political Science
  • Psychology
  • Recreation
  • Social Computing
  • Social Media
  • Social Sciences

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.