Model Concept of a Knowledge System for Advancement in Decision Making

1
Model Concept of a Knowledge System for
Advancement in Decision Making

• Zdenka Winklerova
• University of Defence
• Centre of Advanced Studies
• E-mail zdenka.winklerova_at_unob.cz

2
Motivation

• improvement of decisions by means of
  intelligence interpretation of information
• using knowledge as processed information in
  support of decision making
• not only taking information of all sensors, but
• processing it into useful knowledge
• using this knowledge to choose between
  alternatives while making decisions

3
Leaving Points

• Theory of Knowledge (Epistemology)
• James Frederick Ferrier (1808 -1864).
• Observe Orient Decide Act (OODA) Loop
• Strategic Theory of John Boyd 1
• Network Centric Warfare (NCW) Concept
• Military Information Age Transformation as
  defined by David S. Alberts 2
• JDL Schema
• Data / Information Fusion Model 3

4
Theory of Knowledge

• Data Information Knowledge Wisdom (DIKW)
  epistemological hierarchy

5
OODA Loop
Observe
Orient
Act
Decide
6
NCW Concept

• NCW tenets 2pp.7-8
• a robustly networked force improves information
  sharing
• information sharing enhances the quality of
  information and shared situational awareness
• shared situational awareness enables
  collaboration and self-synchronization, and
  enhances sustainability and speed of command
• these, in turn, dramatically increase mission
  effectiveness

7
JDL Schema

• Multilevel process of Data and Information Fusion
  
• Level 0 Data Producing
• Level 1 Object Refinement
• What is happening at the moment?
• Level 2 Situations
• What does it mean?
• Level 3 Contextual Understanding
• What could happen in the (near) future?
• Level 4 Feedback
• Level 5 Situational Management

8
Statement of the Problem

• Initial Problem Formulation
• Measures of Merit
• Measures of Command and Control Effectiveness
  (MoCE)
• Measures of Command and Control Performance (MoP)
• Independent Variables
• Objective Setting
• Scope and Assumptions

9
Initial Problem Formulation

• Information Management Challenges
• Method of networking the (force) entities
• Course of Information Sharing
• Informational complexity consisting in potential
  overloading of individuals
• Managing the informational and computational
  complexities
• Semantic interoperability
• Collaborative behaviour
• Social coordination of collectives
• Consistency with Mission Capability Package
• concept of operations and mission intent, command
  approach, function of control and force
  organization

10
(No Transcript)
11
Measures of Force Effectiveness

• C2 Maturity Function ( Degree of
  Decentralization Range X,
  Degree of Information Sharing
  Range Y,
  Degree of Social Interaction Range Z ) Range M?
• Range X Cyclic, Interventionist,
  Problem Solving, Problem Bounding,
  Selective Control, Control Free ?
• Range Y No-Sharing, Shared Perception,
  Shared Comprehension,
  Shared Projection ?
• Range Z Non-Interactive, Reactive,
  Interactive, Interconnected ?
• Range M Conflicted, De-conflicted,
  Coordinated, Collaborative, Agile ?

12
Measures of Force Effectiveness (...)

• Degree of Decentralization Function (
  Investigation,
  Level of
  Expertise,
  Situational
  Awareness )?
• Degree of Information Sharing Function (
  Situational Awareness,
  Social
  Awareness )?
• Degree of Social Interaction Function (
  Interaction,
  Coordination,
  
  Integration )?

13
Measures of Performance

• Investigation yes, no ?
• Helping the force entities to find information
  relevant to their current activities quickly
• Situational Awareness None, Classification,
  Perception,
  Comprehension, Projection ?
• Social Awareness yes, no?
• Making force entities aware when something
  relevantto them occurs
• Classification yes, no ?
• Mediating the force entity the Object Refinement
• What is happening at the moment?

14
Measures of Performance (...)

• Perception yes, no ?
• Mediating the force entity the Situations
  Assessment including Classification
• What does it mean?
• Comprehension yes, no ?
• Mediating the force entity the Impact Assessment
  including Perception
• What could happen in the (near) future?
• Projection yes, no?
• Allowing the force entities to identify the
  consequence of the comprehended situations for
  their own intent (projection)
• Allowing the Planning and Scheduling of the
  Feedback for creating of the (shared) command
  intent.

15
Measures of Performance (...)

• Social Awareness yes, no ?
• Making force entities aware when something
  relevantto them occurs
• Interaction yes, no ?
• Making the force entity able to use the
  concurrent work of the other entities while
  working on a shared task
• Coordination yes, no ?
• Getting the force entities to focus their
  activities on the right things at the right time
  
• Integration yes, no ?
• Making the force entity able to combine results
  produced by other entities

16
Input Variables

• Message Representation yes, no ?
• Getting the force entities to understand each
  other through the semantic content of the
  exchanged messages
• Yes / no parameter
• Level of Expertise true, false ?
• The level of expertise of the force entity
  personnel,
• prescribed / other
• True / false parameter

17
Objective Setting

• Concept of Knowledge System
• Utilization of the Information and/or Knowledge
  in course of the C2 cycle
• Monitoring and Understanding Module
• corresponding with Classification Element (Object
  Refinement), Perception Element (Situations),
  and Comprehension Element (Impact Assessment) of
  the Situational Awareness
• Planning and Scheduling Module
• corresponding with the Projection element of the
  Situational Awareness for creating of the
  (shared) command intent (Feedback)
• Execution and Control Module
• implicating the ability of the Social Awareness,
  Investigation, Interaction, Coordination, and
  Integration

18
Scope and Assumptions

• Representing the Community of Interest (COI) as a
  multi agent system
• Applying ontological approach to the
  communication among users within a COI
• Using a social agreement protocol for purpose of
  the social coordination of collaborating entities
• Ability of the agent-based planning and
  scheduling to switch to a new schedule and/or
  plan that was pre-prepared in the meantime (while
  executing the original schedule) whenever the
  original plans are frozen
• Concept of the knowledge system as a common and
  shared information infrastructure that would
  allow any course of command and control

19
Findings

• Model Concept of the resulting Knowledge System
• Monitoring and Understanding Module
• Classification Element
• Perception Element
• Comprehension Element
• Planning and Scheduling Module
• Execution and Control Module

20
Model Concept Basic View
21
Classification
22
Perception
Classification Elements(Objects)
23
Comprehension
24
Planning and Scheduling

• dynamic networking of (battle space) entities
• allowing for use of hierarchies when appropriate
• as well as for the non-hierarchical networks
• entities in Community of Interest (COI)
• multi-agent community
• each (force) entity represented as an intelligent
  social agent
• decentralization of intent
• ability of each agent
• to ask (i.e. pull the awareness), to tell (i.e.
  push the awareness)
• to pull (accept) the intent and also to push the
  intent.
• allowing for agreements (shared projection) about
  intent
• as well as about the awareness
• social model of (military) capability planning
• agent is free to contract non-linearly
• than having to adhere to a linear top down model
  

25
Social Model for Capability Planning
26
Execution and Control

• agent based simulation of external process
• simulation of the real behaviour of the agents
• by community of emulation agents
• organized as an interaction of two emulations
• emulation of the external process
• emulation of agents control algorithm
• the emulation agent(s)
• simulate the behaviour of a real execution of a
  plan
• by modelling the feedback from the plan
  realization
• provide feedback to the Planning and Scheduling
  process
• the real agent
• initiate re-planning when actual course of
  execution differs from the plan

27
Levels 0-3 Monitoring ?UnderstandingModule
Level 4Execution ?ControlModule
Level 4Planning ? SchedulingModule
28
Recommendations

• 3bA agent approach and the multi-agent solution
• at Level 4 of the Knowledge System
• aggregation of different functionalities that
  have previously been distinct
• such as Planning and Scheduling, Coordination,
  negotiation, adaptation and learning
• plug-and-play approach
• agents grouped into different types of
  communities such as teams, coalitions, platforms,
  etc. can freely join and leave the communities
• along with the Blackboard (BB) solutions

29
Future R D

• application of Transparent Intensional Logic
  (TIL)
• for achieving semantic communication and
  information fusion / integration
• (under consideration)
• demand on parallel extension of the classic,
  single-threaded blackboard architecture
• while allowing true concurrent KS executions
• BB as a meta-agent for projection of a common
  intent?

30
Risk Assessment of the Study

• adoption of NATO Code Of Best Practice (COBP)
  methodology
• to minimise the risk of the study
• COBP cannot fully eliminate the risk
• risk associated primarily with
• the problem formulation
• establishing the values of the variables by
  assumptions
• diversity of courses of actions that are possible
  when involving human decision making
• Level of Expertise is treated as an independent
  input variable
• possibility of complex (even chaotic) behaviour
  arising from the dynamic interactions of the
  collaborating entities

31
References

1. Osinga, F. Science, Strategy and War The
   Strategic Theory of John Boyd. ISBN 0-415-37103-1
   . Abingdon, United Kingdom, Routledge, 2007.
2. Alberts, D. S. Information Age Transformation.
   Getting to a 21st Century Military. CCRP
   Publication Series, Washington, D.C., ISBN
   1-893723-06-2, First printing 1999, 2nd printing
   Oct 1996, Revision, 2002.
3. Steinburg, A. N., Bowman, Ch. L., White, F. E.
   Revisions to the JDL Data Fusion Model. In
   Proceedings of the Joint NATO/IRIS Conference,
   Quebec, 1998.
4. NATO Code of Best Practice for Command and
   Control Assessments. CCRP Publication Series,
   Washington, D.C., ISBN 1-893723-09-7, 1st
   printing Oct 2002. Reprint July 2004.
5. Gorodetski, V., Karsayev, O., Samoilov, V.
   Multi-agent Data Fusion Systems Design and
   Implementation Issues. In Proceedings of the 10th
   International Conference on Telecommunication
   Systems Modeling and Analysis, Monterey, CA,
   VOL 2, pp. 762-774, October 3-6, 2002.

32
References (...)

1. Hecking, M. Analysis of Free Form Battlefield
   Reports with Shallow Parsing Techniques. In
   Proceedings of the Military Data and Information
   Fusion Symposium, RTO IST-040 / RSY-012, Prague,
   October 20-22, 2003.
2. Lorenz, F. P., Biermann J. A Man-in-the-Loop
   Support Concept for Military Ambush Threat
   Assessment Based on Reconnaissance Reports. In
   Proceedings of the Military Data and Information
   Fusion Symposium, RTO IST-040 / RSY-012, Prague,
   October 20-22, 2003.
3. Winklerova, Z. Ontological Approach to the
   Representation of Military Knowledge. In
   Proceedings of the Military Data and Information
   Fusion Symposium, RTO-IST-040 / RSY-012.
   Prague, October 20-22, 2003.
4. Pechoucek, M., Marik, V. Social Models and their
   Applications. Research Report. Czech Technical
   University in Prague, Faculty of Electrical
   Engineering, Prague 2004.
5. Marik, V. Conceptual and Technical Problems
   Connected with Expansion of the Agent-Based
   Solutions. Research Report. Czech Technical
   University in Prague, Faculty of Electrical
   Engineering, Prague 2004.

33
Dual Model (suggested)