Ontology-Based M&S Dictionaries:
An Example for V&V
update and ã by Tuncer Ören 2006-09-09
http://www.site.uottawa.ca/~oren/
Ontologies are used to represent relationships of concepts. See for example Fishwick and Miller for ontologies for M&S (2004). There are different ways relationships can be represented. For example in Visual Thesaurus, a graphic relationship is augmented with definitions [VT]. Ontology-based dictionaries can be used to represent the conceptual relationships of M&S terms.
Several interwoven concepts and terms exist related with quality of modeling and simulation, software, user/system interfaces, as well as V&V techniques. Ways to classify and present the relevant information as an ontology-based dictionary would be useful for practitioners as well as for students.
As an example, we elaborate on a small number of M&S V&V terms given in Table 1 [Ören et al., 1985]. For a complete list of M&S quality assurance terms, please see Ören [1984]. Table 2 is a conventional dictionary (arranged alphabetically by the terms) for the M&S V&V terms listed in Table 1.
In quality assurance studies, most of the terms represent evaluation of an element with respect to a criterion. Hence, in an ontology-based dictionary, as in Tables 3 and 4, col. 1 shows the element which is evaluated while col. 3 shows the criterion with respect to which the evaluation is done. Col. 2 can be used to refer to a subset of an element represented in col. 1. In reading a definition of a term in an ontology-based dictionary, one can also see –in a systematic way– the relationship (and the nature of relationship) of a term with other terms. In an ontology-based dictionary, the relationships can be very varied.
For cases where an evaluation criterion cannot be identified, col. 3 can be left blank.
A regular dictionary (arranged alphabetically by the terms) would consist of only columns 4 and 5 to display the terms and the corresponding definitions and possibly pointers to relevant additional concepts.
A 6th column can –be added to a systematic dictionary to indicate pointers to applicable technique(s). Similarly a 7th column can be added for references.
An ontology-based dictionary can be on a CD or can be made available on a Web site where all the searchable terms can be arranged in another alphabetical list. For example, the list of the searchable terms covered in Tables 3 and 4 are shown in Table 1. (Since Tables 3 and 4 are provided as examples –they don’t include all the terms; hence, the terms given in Table 1 are a small subset of the relevant terms.) Terms consisting of more than one word, e.g., “experimentation error” can be listed more than once, e.g., “experimentation error” and “error, experimentation,” to be able to search them grouped in different ways; in this case, under experimentation and error.
References:
Fishwick, P.A., Miller, J.A. (2004). Ontologies for Modeling and Simulation: Issues and Approaches. Proceedings of the Winter Simulation Conference, 259-264.
Ören, T.I. (1984). Quality Assurance in Modelling and Simulation: A Taxonomy. In: Simulation and Model-Based Methodologies: An Integrative View, T.I. Ören, B.P. Zeigler, M.S. Elzas (eds.). Springer-Verlag, Heidelberg, Germany, pp. 477-517.
Ören, T.I., Elzas, M.S., Sheng, G. (1985). Model Reliability and Software Quality Assurance in Simulation of Nuclear Fuel Waste Management Systems. In: Waste Management '85, Vol. 1, R.G. Post (ed.), pp. 381-396.
VT - Visual Thesaurus. http://www.visualthesaurus.com/
Publications of Dr. Tuncer Ören
- on M&S Quality Assurance (QA) are listed at:
http://www.site.uottawa.ca/~oren/pubsList/QA.ht
- on M&S taxonomies are listed at:
http://www.site.uottawa.ca/~oren/pubsList/taxonomies.htm
Table 1. List of Searchable Terms for the Example on
an Ontology-based Dictionary for M&S V&V
Behavior, sensitivity of model |
Behavioral comparison of models |
Behavioral model validity |
Check, formal |
Comparison of models, behavioral |
Comparison of models, structural |
Computerized experiments, software quality assurance of |
Conditions, verification of experimental |
Data relevance |
Error, experimentation |
Error, instrumentation |
Experimental conditions, verification of |
Experimentation error |
Experiments, software quality assurance of computerized |
Formal check |
Instrumentation error |
Library, software quality assurance of run-time simulation |
Model behavior, sensitivity of |
Model qualification |
Model realism |
Model validity, behavioral |
Model verification |
Model verification, structural |
Model, software quality assurance of simulation |
Models, behavioral comparison of |
Models, structural comparison of |
Qualification, model |
Quality assurance of computerized experiments, software |
Quality assurance of run-time simulation library, software |
Quality assurance of simulation model, software |
Realism, model |
Relevance, data |
Run-time simulation library, software quality assurance of |
Sensitivity of model behavior |
Simulation library, software quality assurance of run-time |
Simulation model, software quality assurance of |
Software quality assurance of computerized experiments |
Software quality assurance of run-time simulation library |
Software quality assurance of simulation model |
Structural model comparison |
Structural model verification |
Validity, behavioural model |
Verification of experimental conditions |
Verification, model |
Verification, structural model |
Table 2. A Conventional Dictionary for the M&S V&V Terms Listed in Table 1
Terms |
Definition/Comment (and relevant concepts) |
|
|
Behavioral comparison of models |
Comparison of behavior of a model with the behavior of another model generated under same conditions |
Behavioral model validity (as a black box) |
Comparison of model behavior and real system behavior observed or generated under same/similar conditions. |
Data relevance |
Evaluation of data collected from the real system with respect to the goal of the study. (data accuracy) |
Experimentation error |
Evaluation of the procedure of collecting data from real system with respect to experimental conditions. |
Formal check |
Evaluation of the simulation model with respect to the modeling formalism used. (consistency checks, completeness checks for built-in quality assurance) |
Instrumentation error |
Evaluation of data collected from the real system with respect to real-system as the source of data, taking into account error tolerance. |
Model qualification |
Evaluation of a conceptual model with respect to the goal of the study. (model relevance, model adequacy) |
Model realism |
Evaluation of a conceptual model with respect to the real system. |
Model verification |
Evaluation of a simulation model with respect to a conceptual model |
Sensitivity of model behavior |
Comparison of model behavior under different scenarios where all conditions are kept same except for the parameter and/or initial condition for which sensitivity is being assessed. |
Software quality assurance of computerized experiments |
Evaluation of computerization of the experiments with respect to software quality assurance requirements. |
Software quality assurance of run-time simulation library |
Evaluation of the simulation run-time library with respect to software quality assurance requirements. |
Software quality assurance of simulation model |
Evaluation of a simulation model with respect to software quality assurance requirements. |
Structural model comparison |
Evaluation of the structure of a simulation model with respect to the structure of another simulation model. |
Structural model verification |
Evaluation of the structure of a simulation model with respect to perceived structure of real system. |
Verification of experimental conditions |
Evaluation of a computerization of experiments with respect to experimental conditions |
Table 3. Part of an Ontology-based Dictionary
1
2
3
4
5
Evaluation of
with respect to
Term
Definition/Comment
(and relevant concepts)
Conceptual model
Goal
Model qualification
Evaluation of a conceptual model with respect to the goal of the study. (model relevance, model adequacy)
Real system
Model realism
Evaluation of a conceptual model with respect to the real system.
Data
(behavior):
Real system data
(Real system behavior)
Goal
Data relevance
Evaluation of data collected from the real system with respect to the goal of the study. (data accuracy)
Real system
Instrumenta-tion error
Evaluation of data collected from the real system with respect to real-system as the source of data, taking into account error tolerance.
Experimental conditions
Experimenta-tion error
Evaluation of the procedure of collecting data from real system with respect to experimental conditions.
Simulated data
(Model behavior)
Real system behavior
Behavioral model validity (as a black box)
Comparison of model behavior and real system behavior observed or generated under same/similar conditions.
Behavior of same model under different scenarios
Sensitivity of model behavior
Comparison of model behavior under different scenarios where all conditions are kept same, except for the parameter and/or initial condition for which sensitivity is being assessed.
Behavior of another model under same scenario
Behavioral comparison of models
Comparison of behavior of a model with the behavior of another model generated under same conditions
Table 4 Part of an Ontology-based Dictionary
1
2
3
4
5
Evaluation of
with respect to
Term
Definition/Comment
(and relevant concepts)
Simulation
Program:
Simulation model
(Computerized model)
Real system
Structural model verification
Evaluation of the structure of a model with respect to perceived structure of real system
Conceptual model
Model verification
Evaluation of a simulation model with respect to a conceptual model
Another model
Structural model comparison
Evaluation of the structure of a simulation model with respect to the structure of another simulation model.
Modelling formalism
Formal check
Evaluation of the simulation model with respect to the modeling formalism used. (consistency checks, completeness checks for built-in quality assurance)
Software engineering requirements
Software quality assurance of simulation model
Evaluation of a simulation model with respect to software quality assurance requirements.
Computerized experiments
Experimental conditions
Verification of experimental conditions
Evaluation of a computerization of experiments with respect to experimental conditions
Software engineering requirements
Software quality assurance of computerized experiments
Evaluation of computerization of the experiments with respect to software quality assurance requirements.
Run-time simulation library
Software engineering requirements
Software quality assurance of run-time simulation library
Evaluation of the simulation run-time library with respect to software quality assurance requirements.