Situational Behavior Modeling

Behavior modeling for military applications needs to consider systems in which all kind of entities participate - machines, humans, human organizations (like platoons or companies) as well as such complex entities like countries, industries and societies. The variety and the structure of entities pa...

Full description

Saved in:
Bibliographic Details
Main Authors: Matheus, Christopher J, Ulicny, Brian, Kokar, Mieczyslaw M, Baclaawski, Kenneth, Ng, Won, Lau, Lena, Letkowski, Jerzy, Weyman, Jerzy, Dionne, Robert, Parent, Douglas
Format: Report
Language:English
Subjects:
BAYES THEOREM
BAYESIAN NETWORKS
BEHAVIOR
DYNAMIC MODELS
GAME THEORY
HUMANS
INFORMATION PROCESSING
Information Science
MARKOV PROCESSES
MATHEMATICAL MODELS
MILITARY APPLICATIONS
MILITARY REQUIREMENTS
NETWORKS
ONTOLOGY
Psychology
SITUATION TRACKING
SITUATIONAL AWARENESS
SITUATIONAL BEHAVIOR
Statistics and Probability
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Behavior modeling for military applications needs to consider systems in which all kind of entities participate - machines, humans, human organizations (like platoons or companies) as well as such complex entities like countries, industries and societies. The variety and the structure of entities participating in behaviors in the military domain require the use of representations and tools appropriate for this kind of complexity. Ontological modeling seems to be the best match for this domain. However, there are no known results in the literature on modeling and tracking of behaviors using an ontological approach in which automatic inference over the dynamic models of behaviors can be carried out using inference tools. A behavior model can be conceptualized in a number of ways - as an abstract concept that is independent of any physical or conceptual entity, as a feature of a specific entity, or as an abstract concept that is associated with one or more physical or conceptual entities. Various knowledge representation mechanisms including State Machines, Hidden Markov Models, Petri Nets, Game Theoretic Models and Bayesian Networks have been used extensively for behavior modeling. Most of the studies have been focusing on modeling behavior of a specific type of entity. For instance, organizational behavior modeling considers an organization as a system of interrelated entities (humans) and then develops models for behavior of humans within an organization. In the approach presented in this document, behavior is treated as being associated with a situation, i.e., with a number of objects (e.g., an organization) being in some relations with each other. While situation objects will normally have some basic behaviors associated by default, they will be able to participate in complex behaviors involving multiple situation objects.