Page Last Updated: EDT, © 2010, 2011, 2016

HARTLEY CONSULTING
Solving Complex Operational and Organizational Problems

PROJECT: IW Metric Ontology

Dr. Dean S. Hartley III

Project Metadata Keywords
Label Name Other Year DurationYrs
Client TRAC US Army
Dates 2010 0.75
Employer Hartley Consulting
Partner Dynamics Research Corporation (DRC)
Pubs "Chapter 2: Creating the foundations for modeling irregular warfare," in Advances in Design for Cross-Cultural Activities, Part II, Eds. Denise M. Nicholson and Dylan D. Schmorrow, CRC Press, Boca Raton, FL lead author 2012
Pubs "Ontology Structures for Modeling Irregular Warfare," in Military Operations Research, Vol17, Num 2, 2012 author 2012
Pubs “Developing Human Social Cultural and Behavioral (HSCB) Ontologies to Support Simulations," Simulation Interoperability Workshop 2011, April 2011, Orlando, FL co-author 2011
Pubs "Using Ontologies in Conceptual Model Validation," in Hawaii University International Conferences (HUIC) Proceedings, July 2012 author 2012
Pubs Irregular Warfare (IW) Metrics Ontology Final Report. Dynamics Research Corporation, Orlando, FL lead author 2011
Team Lee W. Lacy
Configuration management
Consequence Management
Data Verification & Validation
Database design
DIME/PMESII Modeling
Documentation standards
Geopolitical analysis
Global War on Terrorism (GWOT)
Human factors
Human, Social, Cultural Behavior (HSCB) Modeling
Impact analysis
Independent Verification & Validation (IV&V)
Information storage and retrieval
Irregular Warfare (IW)
Knowledge Management (KM)
Metadata
Model/System integration
Modeling, Simulation & Gaming (MSG)
Ontologies
Operations Other Than War (OOTW)
Software issues
Software reuse
Stability Operations (SASO, SSTR)
Verification, Validation & Accreditation (VV&A)
Warfare modeling

Challenge:

Support the US Army Training and Doctrine Command (TRADOC) Analysis Center (TRAC) in developing an ontology of the metrics needed for modeling Irregular Warfare (IW).

Background:

The initiative is led by the TRADOC Analysis Center (TRAC). The Methods, Models and Tools (MMT) aspect of the TRAC IW Analytic Capability (IWAC) program is primarily focused on the Irregular Warfare Tactical Wargame (IW TWG). The IW TWG is implemented as a composition of tool modules. Some of the input data to the TWG describe the state of the operational environment and some of the outputs describe perceptions of the population and key individuals. The Army envisions its mission as supporting one or more Lines of Effort (LOEs), with intermediate metrics in the PMESII spectrum to be used as indicators of success.  The goal is to provide results that indicate perceptions of key personnel and population, describe the current state of the operational environment, and to determine progress towards the desired end state. The IWAC program should support multiple echelons (e.g., company, battalion).

LOEs and the DIME/PMESII Paradigm:

The six LOEs within the current scope are the first six in the table to the right, taken from FM3-24.2, pg 4-8, 21 April 09.  Each LOE may be thought of as a large task that  can be broken into subtasks, several of which.are included in the table.  Each LOE also has a desired endstate - which is not a task or action, but the value of a state variable.  It should be noted that these are Department of Defense LOEs and that other parts of the government, the parts of the relevant host nation (HN), and other interested parties will have their own analogs to these LOEs.

PMESII variables are state variables that describe the state of the world by dividing it into Political, Military, Economic, Social, Information, and Infrastructure categories.  These categories are further subdivided for convenience.  Each of the LOE endstates can be placed somewhere within the set of PMESII variables.

DIME variables represent actions that can be undertaken by a government and are divided into Diplomatic, Informational, Military and Economic Categories.  DIME actions are generally at the level of sub-subtasks as compared to this table.


The figure to the right illustrates one possible model connecting DIME actions (red-colored boxes) and PMESII state variables (other boxes).  Additional annotations have been added to show the connections to one LOE.

LOE 5 is a task to provide support to economic and infrastructure development. One DIME action that might be employed in this task is rebuilding roads (point #1 at the top in the figure and the circled red box on the right edge of the figure).  This model claims that this action will result in some amount of road capacity being created, some amount of jobs being created, and some amount of investment (point #2 and the other circled boxes on the lower right).  These variables will in turn lead to modified HN road capacity, jobs and foreign investment (point #3 and magenta circled boxes in the interior of the figure).  If the theory and model are correct, eventually the connections will lead to improvement in the LOE 5 Endstate variable (blue circle at the top of the figure).  Following the model connections will also lead to other consequences, some beneficial and perhaps some harmful.

The figure also shows that the model connections can be followed in reverse (points #1 & 2 at the bottom left and the three blue circled variables).  Tracking backward from the endstate variable will identify all of the actions that could affect the endstate.  It may also identify variables that are not under the control of a particular side that could affect the endstate.

 

Process:

Dynamics Research Corporation (DRC) and Hartley Consulting created a team to address the problem, with DRC supplying the ontology expertise and Hartley Consulting supplying the IW modeling expertise.

As shown in the figure to the right, the goal of producing an IW ontology and linking the LOEs to PMESII Metrics was envisioned as being created through
  • Workshops,
  • Literature review,
  • Tools review, and
  • Expertise.

Literature review included a search for general concepts and relationships from doctrine (such as Joint Publications and Army Field Manuals), workshops (including those hosted by the Military Operations Research Society (MORS), the National Defense University (NDU), and the Human Social Cultural Behavior (HSCB) project), and books articles and presentations.

In this process the team would develop

  • IW definitions,
  • Ontology definitions,
  • LOE definitions,
  • PMESII definitions,
  • Metric definitions, and
  • Operational knowledge.

The results would be collected into a final report that would describe the machine readable ontology.


The ontology creation process is illustrated by the IDEF0 diagram to the right. 

Step A1 consists of scoping the problem domain, based on the tasking.

Step A2 consists of identifying ontologies that can be reused, based on existing ontologies and controlled by the ontology requirements output in A1.

Step A3 consists of defining the classes and properties needed in the IW ontology, based on the ontology reuse candidates output by A2 (and the ontology requirements). 

Step A4 consists of documenting and encoding the ontology design artifacts.

 

Ontology Background:

An ontology is a tool used in describing a domain of knowledge about the world.  Accordingly, it consists of vocabularies, definitions and relations.  The figure to the right, derived from Uschold, shows a spectrum of increasingly formal descriptions of a domain, in this case illustrated by a "pump."

The figure to the right, taken from McGuinness, shows a spectrum of descriptive techniques, also ranging from informal to increasingly formal, with a dividing line indicating the general area in which a description can be called an ontology.  The "is-a" relationship provides a separation point.  Stating that a pie is a type of dessert is an informal "is-a" relationship. A formal "is-a" relationship requires that the vocabulary is controlled so that "desserts" are distinguished from non-desserts, "pies" from "cakes," etc.

 

The figure to the right begins the process of describing this formalization.  The domain (for us IW metrics) is represented by a world globe. The domain is modeled by a conceptualization that consists of a controlled vocabulary of classes and their relationships.  The ontology specifies this conceptualization (for instance by using a computer readable language) and thus describes the domain.

The ontology is divided into two parts, the Tbox elements (class names, attributes, and relations) and the Abox elements (instances of the classes and qualifiers).  As examples: "pie" could be a class, with attributes such as definition and definition author; "peach pie" could be a class, related to "pie" by the "is-a" relation, with attributes such as "must contain peaches or peach flavoring"; and a particular peach pie could be an instance, with qualifiers such as "sell-by-date."

 

The most formal types of ontologies are encoded in some machine readable language to permit automatic inferencing and use by general programs.  The figure to the right, derived from Lacy, shows how an application (such as the IW Metric Ontology) is based on a high level language, in this case the Web Ontology Language (OWL) 2.  OWL 2, in turn is based on lower level languages, down to XML (which is a generalization of the HTML used in web pages) and IRIs (generalizations of URLs, the location pointers for web pages).

 

Technical Results:

The central organizational principle for the development of the IW Metric Ontology is provided by the figure to the right. The Operational Environment that includes everything relevant to irregular warfare is divided into three parts: Actors, Actions, and the Environment.

Actors are human and natural entities that cause things to happen, thereby changing things.

Actions are the interventions, events, and ongoing processes that are performed by actors and which directly cause changes.

The Environment represents the rest of entities in the Operational Environment.

Actors perform Actions, which affect the Operational Environment (OE). The state of OE, including any changes, is described by State Variables. Actors perceive the OE by means of the State Variables.

State Variables include both numeric variables (true metrics) and categorical variables (e.g., type of government).  IW Metrics are defined to be these State Variables and provide the content of the IW Metric Ontology.


The formative taxonomy that was chosen was the PMESII framework because its content includes the IW State Variables.  The initial sources for the metrics were Hayes & Sands Doing Windows, the Interim Semi-static Stability Model (ISSM), and the DIME/PMESII VV&A Tool. Additional sources provided confirmation of the metrics and additional metrics.  The major additional sources were the Klein's HSCB Taxonomy, TRACs Metrics v3, the OCRS Matrix, Hartley's Corruption in Afghanistan, TRAC's IW Decomposition, Dziedzic's MPICE, SRI's PRIME Taxonomy, and Hilson's Requirements document (citations below in the Literature section).

The figure to the right displays the initial taxonomic organization.  The actual State Variables are attached to the subcategories. Although this taxonomy has been revised (so that IW Metrics relating to Actors and Actions [in particular the direct results of DIME actions] are included), note that even in its early state, this taxonomy extends beyond a pure PMESII taxonomy: the Kinetic Evironment and the Natural Environment have been added.  Technically, this is a PMESII+ taxonomy; however, for brevity, it is often referred to as a PMESII taxonomy.


One of the differences between an ontology and a taxonomy is structural and conceptual. Each element in a taxonomy can only have a single "parent," whereas in an ontology, an element may have multiple "parents."  For example, in a taxonomy, buildings used by a government can be placed in Government under Infrastructure Environment or in Government under Economic Environment or in Government under Military Environment or in some other part of the structure, but not in more than one part. In an ontology, these buildings may be connected to multiple "parents," as appropriate. For example, government educational buildings would be connected to both the Education box under Infrastructure Environment and to the Education box under Social Environment. This means that the attributes for the class of educational buildings can be attached to a single class and the qualifiers for individual educational buildings only need to be entered in one place, yet they can be retrieved through either route.

In addition to linking the IW Metrics to multiple PMESII categories, the ontology structure allows linking the metrics to a completely different structure - the LOE structure, described in the Background above. Thus each direct result of a DIME Action can be linked to the appropriate LOE or LOEs. Further, the other metrics can also be linked to the LOEs.  This supports the goal of providing information connecting the state of the OE to progress toward achieving the desired endstate or endstates.

Another difference between ontologies and taxonomies is useful in making the creation process easier. Previously defined ontologies can be included, reducing the need to create structures. For example, the Dublin Core Ontology, partially illustrated in the figure to the right, provides a structure for defining the Attributes of the classes and the Qualifiers of the instances. The IES Ontology, created for TRAC by another project provides another example of connecting ontologies.  The IES Ontology doesn't provide structures that the IW Metric Ontology needs; however, it provides additional content covering general infrastructure.  Because it is an ontology, it can be linked to the IW Metric Ontology, increasing the value of both ontologies.


During the creation of the ontology, the team used a set of Excel spreadsheet to collect and organize the information as it was created (steps A1, A2, and A3).

In step A4, the spreadsheets were first converted to a database, making explicit the multiple parent relationships that could only be represented by text comments in the spreadsheets. The database was converted to OWL ontology files using a DRC tool (the DOAT Database tool), as illustrated in the figure to the right.

The logical structure of the OWL files is illustrated in the figure to the right. The gray block on the left represents the PMESII+ structure (in part). The superclass OperationalEnviornmentCharacteristics is a subclass of the Metrics class and has the various PMESII categories and subcategories as subclasses, so they inherit the attributes of a general metric. The actual metrics (such as FreedomRating) are subclasses of the general metric class, as well as being subclasses of the appropriate subcategory class(es). The metrics are also properties of the appropriate LOEs. Each metric also has as subclasses the source-data classes, which have instance data.

 

The figure to the right illustrates the physical structure of the OWL ontology files. This physical modularization makes it easier for application developers to import only the portions of the ontology that they need and separates common information into single files.

 

Operational Results:

During the project workshops, TRAC used the ontology in several ways.

  1. Army personnel examined each metric element for connections to the LOEs.
    • This process generated suspected primary, secondary and tertiary connections between each element and the set of LOEs.
    • The process also provided subject matter validation of the metric elements.
    • And the process familiarized the TRAC team with the details of the ontology.
  2. Army personnel examined each metric element for utility concerning the IW TWG.
    • The first question was, "is this element already addressed by the TWG?"
    • The second question was, "if not, should it be added to next year's improvements list?"
    • The third question was, "if not, should it be added to the 'wish list'?"
  3. Army personnel created analogs of the LOES for participants other than DoD (e.g., the Host Nation, the Department of State, the adversary) and connected them to the metric elements.

Conclusions:

Developing the IW Metrics ontology resulted in a series of technical conclusions.  These include:

 Additionally, there are communications and other benefits including:

 

Recommendations:

The team advised TRAC that several steps can be taken to beneficially extend the results.

Identify and document use cases, such as the following:

Perform additional ontology work, such as the following:

This additional ontology work was performed as an IR&D project in Total IW Ontology.

Literature:

Doctrine

US Government COIN Guide
IW Joint Operating Concept
JP 3-24
Joint Doctrine Pub 3-40
FM 3-0
FM 3-24
FM 3-07
FM 3-07.1
FM 3-24.2
CALL Leader’s Handbook 07-27
CALL PRT Playbook

Lexicon

JP 1-02, DoD Dictionary of Military and Associated Terms
Some terms identified as part of Capabilities Based Planning
Some terms identified as part of MORS workshops
Other terms will be defined as part of ontology development effort

Subject Matter - Books

Davies, J. L. & Gurr, T. R., (Eds.) (1998). Preventive Measures: Building Risk Assessment and Crisis Early Warning Systems. Lanham, MD: Rowan & Littlefield.
Galula, D. (1964). Counterinsurgency Warfare: Theory and Practice. Westport, CN: Praeger Secutity International.
Gilbert, N. & Troitzsch, K. G. (2005). Simulation for the Social Scientist (2nd ed).Glasgow, U.K.: Bell & Bain.
Grady, R. B., & Caswell, D. L. (1987). Software Metrics: Establishing a Company-Wide Program. Englewood Cliffs, NJ: Prentice-Hall.
Hayes, B. C., & Sands, J. I. (1998). Doing Windows: Non-Traditional Military Responses to Complex Emergencies. Washington D.C.: CCRP.
Kaner, S., Lind, L., Toldi, C., Fisk, S., & Berger, D. (2007). Facilitator’s Guide to Participatory Decision-Making (2nd ed.). San Francisco: Jossey-Bass.
Kilcullen, D. (2010). Counterinsurgency. New York: Oxford.
Nelson, R. B., & Wallick, J. (1994). The Presentation Primer: Getting Your Point Across. New York; Irwin.
Orr, R. C. (Ed) (2004). Winning the Peace: An American Strategy for Post-Conflict Reconstruction. Washington, D.C.: Center for Strategic and International Studies.
Rumbaugh, J., Blaha, M., Premerlani, W., Eddy, F., & Lorensen, W. (1991). Object-Oriented Modeling and Design. Englewood Cliffs, NJ: Prentice Hall.
Salmoni, B. A. & Holmes-Eber, P. (2008). Operational Culture for the Warfighter: Principles and Applications. Quantico, VA: Marine Corps University.
Schwarz, R., (2002). The Skilled Facilitator New & Revised: A Comprehensive Resource for Consultants, Facilitators, Managers, Trainers, and Coaches. San Francisco: Jossey-Bass.
Sharp, E. B. (Ed.) (1999). Culture Wars & Local Politics. Lawrence, KS: University Press of Kansas.
Wilkinson, M. (2004). The Secrets of Facilitation: the S.M.A.R.T. Guide to Getting Results with Groups. San Francisco: Jossey-Bass.
Zeigler, B.P., Praehofer, H., & Kim, T. G. (2000). Theory of Modeling and Simulation (2nd ed.). New Delhi, India: Harcourt India.

Subject Matter - Papers and Presentations

Alberts, D. S., & Hayes, R. E. (2003). Power to the Edge: Command…Control… in the information Age. Washington D.C.: Command and Control Research Program.
Bachman, J. A., & Harper, K. A. (Nov 2007). A Toolkit for Building Hybrid, Multi-Resolution PMESII Models. Rome, NY: Air Force Research Laboratory.
Canada (2010). Canada’s Engagement in Afghanistan: Quarterly Report to Parliament For the Period of April 1 to June 30, 2010. Canada: Government of Canada.
Challand, LT COL T. (2009). Tipping Sacred Cows: Moral Potential Through Operational Art. Military Review, Sept-Oct 2009, pp. 19-28.
Jack (2009, 5 Aug). Analyzing Populations in Stability Operations using Cultural Geography. Monterey, CA: U.S. Army TRADOC Analysis Center.
Kilcullen, D. (2007). Counterinsurgency in Iraq: Theory and Practice. A seminar at the Marine Corps Base at Quantico, VA, Sept 26, 2007.
Mandrick, MAJ W. S. (2008). The Ontology of Counterinsurgency. Paper presented at the annual meeting of the MPSA Annual National Conference, Chicago, IL, Apr 03, 2008. http://www.allacademic.com/meta/p267785_index.htm
Mansoor, COL P. R., & Ulrich, MAJ M. S. (Oct 2007). Linking Doctrine to Action: A New Coin Center-of-Gravity Analysis. Fort Leavenworth, KS: Army Combined Arms Center, Army & Marine CounterInsurgency Center.
McCrabb, Maris. (2001). Effects-Based Operations: An Overview. 52 slides. Available online at: http://www.dtic.mil/jointvision/ideas_concepts/ebo.ppt or under ‘Internet Resources’ at http://www.au.af.mil/au/aul/bibs/ebo.htm a January 2008 Maxwell AFB, AL, site.
Smith, J. R., Young, W. C., et. al. (2009, Aug 4). Requirements for a Government Owned DIME/PMESII Model Suite. Presentation.

Taxonomies and Metrics

Abdollahian, Mark, Jacek Kugler, Brice Nicholson, and Hana Oh, "Politics and Power," Estimating Impact: A Handbook of Computational Methods and Models for Anticipating Economic, Social, Political and Security Effects in International Interventions, A. Kott and G Citrenbaum, eds. Springer, New York. 2010.
Bennett, William H.  "Media and Influence," Estimating Impact: A Handbook of Computational Methods and Models for Anticipating Economic, Social, Political and Security Effects in International Interventions, A. Kott and G Citrenbaum, eds. Springer, New York. 2010.
Bhavnani, Ravi,  Dan Miodownik, and Rick Riolo, "Groups and Violence," Estimating Impact: A Handbook of Computational Methods and Models for Anticipating Economic, Social, Political and Security Effects in International Interventions, A. Kott and G Citrenbaum, eds. Springer, New York. 2010.
Duong, Deborah, Robert Turner, and Karl Selke, "Crime and Corruption," Estimating Impact: A Handbook of Computational Methods and Models for Anticipating Economic, Social, Political and Security Effects in International Interventions, A. Kott and G Citrenbaum, eds. Springer, New York. 2010.
Dziedzic, Michael, Barbara Sotirin, and John Agoglia, Measuring Progress in Conflict Environments (MPICE): A Metrics Framework for Assessing Conflict Transformation and Stabilization, Version 1.0. US Institute for Peace, Washington, DC. 2008.
Hayes, Bradd C. and Jeffrey I. Sands, Doing Windows: Non-Traditional Military Responses to Complex Emergencies. CCRP, Washington, DC. 1998.
Hartley III, Dean S. Operations Other Than War (OOTW) Flexible Asymmetric Simulation Technologies (FAST) Prototype Toolbox: ISSM v4.00 Analysts' Guide. DRC, Orlando, FL. 2006.
Hartley III, Dean S. DIME/PMESII VV&A Tool (Software). Hartley Consulting, Oak Ridge, TN. 2009.
Hartley III, Dean S. "Corruption in Afghanistan: Conceptual Model," 21 August 2010
Haskins, Casey  "A Practical Approach to Cultural Insight," Military Review, Sept-Oct 2010.
Hilson, Roger et al., Requirements for a Government Owned DIME/PMESII Model Suite. Office of the Secretary of Defense Modeling & Simulation Steering Committee, Washington, DC. 2009.
Jonker, David and William Wright, "Visualization and Comprehension," Estimating Impact: A Handbook of Computational Methods and Models for Anticipating Economic, Social, Political and Security Effects in International Interventions, A. Kott and G Citrenbaum, eds. Springer, New York. 2010.
Kilcullen, David  Counterinsurgency. Oxford University Press, New York, NY. 2010.
Klein, Gary, HSCB Taxonomy, Mitre
Kott, Alexander and Bruce Skarin, "Insurgency and Security," Estimating Impact: A Handbook of Computational Methods and Models for Anticipating Economic, Social, Political and Security Effects in International Interventions, A. Kott and G Citrenbaum, eds. Springer, New York. 2010.
Lofdahl, Corey  "Governance and Society," Estimating Impact: A Handbook of Computational Methods and Models for Anticipating Economic, Social, Political and Security Effects in International Interventions, A. Kott and G Citrenbaum, eds. Springer, New York. 2010.
Office of the Coordinator for Reconstruction and Stabilization, "Post-Conflict Reconstruction Essential Tasks." US Dept of State, Washington, DC. 2005. http://www.crs.state.gov/index.cfm?fuseaction=public.display&id=10234c2e-a5fc-4333-bd82-037d1d42b725
SRI, PRIME Taxonomy
TRAC, IW Decomposition Analytic Strategy, Overview Briefing for IW WG, 6 January 2009
TRAC, Metrics v3.xls
Young, William C. and Jerry R. Smith, "Requirements for Modeling DIME Actions and PMESII Effects" presented at FOCUS 2010 Conference. 2009.
 

Ontology Technology

Allemang, D., & Hendler, J. (2008). Semantic Web for the Working Ontologist: Effective Modeling in RDFS and OWL. Burlington, MA: Elsevier.
Baader, F., Calvanese, D., McGuinness, D. L., Nardi, D., & Patel-Schneider, P. F. (Eds) (2010). The Description Logic Handbook: Theory, Implementation and Applications (2nd ed.). Cambridge, UK: Cambridge University.
Daconta, M. C. (2007). Information As Product: How to Deliver the Right Information, To the Right Person, At the Right Time. Denver, CO: Outskirts.
Fensel, D., Hendler, J., Lieberman, H., & Wahlster, W. (Eds) (2003). Spinning the Semantic Web. Cambridge, MA: MIT.
Lacy, L. W. (2005) Owl: Representing Information Using the Web Ontology Language. Victoria, B.C: Trafford
Uschold, M. (2003). Where are the Semantics in the Semantic Web? Artificial Intelligence, 24(3), 25-36.

Sharing Results:

The team presented its work in several venues during and after the project, as shown below.

"LOEs and the TRAC IW Metric Ontology," MORS Symposium, June 2011, Monterey, CA.
"Developing Human Social Cultural and Behavioral (HSCB) Ontologies to Support Simulations," Simulation Interoperability Workshop 2011, April 2011, Orlando, FL .
"Irregular Warfare (IW) Metrics Ontology," HSCB Focus 2011, February 2011, Washington, DC.
"IW Ontologies," INFORMS Annual Meeting, Charlotte, NC, 13-16 Nov 2011.

 

If you arrived here using a keyword shortcut, you may use your browser's "back" key to return to the keyword distribution page.

Return to Hartley's Projects Page