THE MOT+ VISUAL LANGUAGE FOR KNOWLEDGEBASED INSTRUCTIONAL DESIGN
By Gilbert Paquette, Michel Léonard, Karin Lundgren-Cayrol
LICEF-CIRTA Research Center and CICE Research Chair
Télé-université, gpaquett@licef.teluq.uquebec.ca
Abstract. This chapter states and explains that a Learning Design is the result of a knowledge
engineering process where knowledge and competencies, learning design, media and delivery models are
constructed in an integrated framework. Consequently, we present our MOT+ general graphical language
and editor that help construct structured interrelated visual models. The MOT+LD editor is the newly added
specialization of this editor for learning designs, producing IMS-LD compliant Units of Learning. The
MOT+OWL editor is another specialization of the general visual language for knowledge and competency
models based on the OWL specification. We situate both models within our taxonomy of knowledge
models respectively as a multi-actor collaborative process and a domain theory. The association between
these “content” models and learning design components is seen as the essential task in an instructional
design methodology, to guide the construction of high quality learning environments.
Keywords. Learning Design, Instructional Engineering, Knowledge Modeling, Visual Language,
Pedagogical Model, Unit of Learning, Ontology Web Language, Representation Language, Prerequisite
Competency, Target Competency, Instructional Model, Delivery Model.
INTRODUCTION.
Building high quality learning designs is a very important and demanding task. It is also a
difficult task that we started to address already a decade ago by progressively building an
instructional engineering method (Paquette et al. 1994, 2005a; Paquette 2003), a delivery system
(Paquette et al, 2005b) and a graphical knowledge modeling editor (Paquette 1996, 2002).
In this on-going work and for the present discussion, the point of view is taken that a Learning
Design is the result of a knowledge engineering process, where knowledge and competencies,
learning design, media and delivery models are constructed in an integrated framework. In the
first section of this chapter, we present the MISAi instructional design method based on these
four models and their relationships to each other. The second section presents the MOT
(Modeling with Object Types) visual language and the specialized editing tools that have been
used in numerous applications. We summarize the theoretical basis of the language, its syntax
and semantic, moreover examples within the MISA instructional design method will be
presented.
The third and fourth sections address the standardization issues and how the MOT+ software is
adapted to provide visual aid to designers building knowledge and/or pedagogical models. The
third section focuses on the learning design models, the IMS-LD specification and the
specialized MOT+LD editor that helps designers build IMS-LD compliant and interoperable
units of learning. The fourth section presents the Ontology Web Language (OWL) and the
specialized MOT+OWL visual editor. We use it to represent domain knowledge models and
target competency that can be used to plan, support staff roles and evaluate the quality of
learning designs. In the fifth section we discuss the association between LD models and OWL
1
models to support what we believe is the central task for knowledge-based instructional design
aiming to support learning environments within the Semantic Web.
Finally, the concluding section will summarize the properties of representation languages that we
have found most useful while designing and using the various specializations of the MOT+
software through its evolution from a general knowledge modeling tool to a standardized tool at
the heart of the instructional design methodology.
INSTRUCTIONAL DESIGN BASED ON VISUAL MODELING.
In this section, we present a synthesis of the main MISA 4.0 Instructional Engineering Method
components and concepts. A knowledge modeling approach using the MOT editor was used to
define the Instructional Engineering method itself, its concepts, processes and principles. And
thus, this method can also be seen as a visual modeling application.
This R&D initiative, started in 1992, has led to the MISA 4.0 version (Paquette 2001a, 2002a)
and to its support tool, called ADISAii (Paquette et al 2001). The editor MOT+ is embedded in
the ADISA system and accessible through a web browser from workstations linked to the
Internet. It can also be used without ADISA together with forms provided by the MISA
documentation. Since 2001, the method has been adapted to the huge standardization work that
has occurred in the eLearning sector; we will address this aspect in later sections of this chapter.
Overview of the Method.
The MISA Learning Engineering process produces specifications of learning environment
grouped in documents called Documentation Elements (DE). Table 1 presents these DEs..
Table 1- MISA 4.0 Documentation Elements – Phases and Axes
Phase 1Definition
100 Organization’s Training System 102 Training Objectives
106 Present Situation
108 Reference Documents
104 Learners’ properties
Knowledge Axis
Pedagogy Axis
Media Axis
Delivery Axis
Phase 2 – Initial
solution
210 Knowledge Model
Orientation Principles
212 Knowledge Model
214 Target
Competencies
230 Media
Principles
240 Delivery
Principles
242 Cost-Benefit
Analysis
Phase 3 – LE
architecture
310 Learning Unit
Content
330 Development
Infrastructure
340 Delivery
Planning
Phase 4 – LE
detailed Design
410 Learning Resource
Content
220 Instructional
Principles
222 Learning
Event Network
224 Learning Unit
Properties
320 Learning
Scenarios
322 Activity
Properties
420 Learning
Resource
Properties
440 Delivery Models
442 Actors and their
resources
444 Tools and
Telecommunication
446 Delivery
Services
Phase 5 Validation
Phase 6 –
Delivery Plan
540 Test Planning
430 Learning
Resource List
432 Learning
Resource Models
434 Media
Elements
436 Source Doc.
542 Revision Decision Log
620 Actors and
Group
Management
630 Learning
System/Resource
Management
640
Maintenance/Quality
Management
610
Knowledge/Competency
Management
2
Each DE results from tasks distributed into 6 phases. Within phase 2, 3, 4 and 6, these DE can
also be viewed according to four axes or dimensions of an eLearning environment: Knowledge,
Pedagogy, Media and Delivery. Presently, MISA 4.0 comprises 35 basic sub-tasks, each
producing one DE, numbered, as shown in table 1, from 100 to 640. The first digit denotes the
phase, the second, the axis, and the third, the sequence number within the axis. A DE is either a
visual model, identified in bold italic in table 1, or a text-based form describing guidelines for a
model or properties of objects in the model.
A Problem Solving Approach in 6 Phases .
MISA proposes a problem solving approach. Each MISA phase is subdivided into a number of
steps where parts of a learning environment or system are constructed. These phases are
sequential, but spiral, with frequent returns to modify the result or previous tasks:
Phase 1: Designers build a description of the training problem, its context and constraints. The
general goal that the solution must fulfill and the main characteristics of the target population are
the most important aspects to address at this point.
Phase 2: Designers define a preliminary training solution, centered on a knowledge model for the
learning domain. Prerequisite and target competencies are associated to the most important
knowledge entities in the model. In this phase, designers also build a first pedagogical visual
model called “the learning event network” grouping the main modules or learning units, their
sequencing and the resources needed to perform them or to be produced by learners and
facilitators.
Phase 3: Designers construct a detailed learning design and specify the infrastructure necessary.
Visual learning scenarios are built for each learning unit defined in phase 2, describing the
learning and facilitating activities, the actors that perform them and the resources needed or
produced by these actors. At the same time, a sub-model of the phase 2 knowledge model is
associated to each learning unit thus defining “the learning unit content”. According to the
evolution of the design, media and delivery principles are refined to prepare the next phase.
Phase 4: Centered on the learning resources and delivery models and the properties of objects in
these models several professionals may work together, content experts, instructional designers
and media designers. Another important concurrent task is the description of the properties of
resources in learning scenarios and the association of a sub-model of the knowledge model to
provide a specification of the “learning resource content”.
Phase 5: The project manager plans the validation of the learning environment and produces a
list of possible revisions and decisions about how to improve the specifications created in the
previous phases.
Phase 6: Designers and project manager prepare elements necessary to the delivery of the
learning environment. It produces a synthetic and global description of the learning environment
for its maintenance and quality management by various actors.
A visual modeling approach.
In each of phases 2, 3 and 4, MISA also proposes the development of the learning environment
along four axes: knowledge and competency (content model), instructional, resources and
delivery. The central product of each axis is one or more visual models.
3
The Knowledge Model centers on a graphical representation of the learning environment content
domain. In this model, the domain’s facts, concepts, procedures and principles are displayed and
interrelated with precise links. Then target and prerequisite competencies are linked to
knowledge element in the model, thus identifying prerequisites and learning objectives for the
Pedagogical Model. Subsequently, knowledge units and competencies are also associated to
learning units and to the resources present in the learning units’ scenario models.
The Instructional Model is essentially a visual network of learning events and units, to which
knowledge and target competencies are associated. Each learning unit is also described by a
visual learning scenario specifying learning and support activities linked to resources in the
environment. Resources holding content (as opposed to tools and services) are associated with a
subset in the knowledge model.
The Learning Resource Models are useful to describe materials (or learning objects) to be adapted
and produced, their media components, source documents and presentation principles as well as
other properties aimed at graphical designers and learning material producers.
Finally, Delivery Models are produced to show how and where actors use or provide learning
materials and resources such as tools, communication means, services and locations, used in the
learning environment. Each Delivery Model is a multi-user workflow, where actors use or
produce resources, while assuming different roles. These processes address organizational issues,
such as group organization, staff assignments, technical help, resource delivery, and so on, which
must be prepared to ensure smooth deployment of a network-based or a distance learning
environment.
Each and every one of these models is built using the MOT+ knowledge representation
technique and tool (Paquette 1999, 2002b). Graphical visual models are the basic DE in each
axis, the backbone of the MISA method. Most of the other tasks, in MISA, describe properties of
objects in these models (e.g., competencies, learning units, resources, roles) as well as their
relationships.
MOT+: A GENERIC VISUAL LANGUAGE AND TOOL.
When designers start building a Learning Environment, two basic questions arise: “Which
knowledge must be acquired, what are the target competencies or educational objectives for that
knowledge?” and “How should the activities and the resources be organized to best achieve
knowledge and competency acquisition?” To help designers solve this type of questions, we
have developed a graphical knowledge modeling method and tools, thus visualizing activity
sequences, actors and tools. In this section, we present the MOT modeling language that serves
that purpose and the MOT+ visual modeling editor.
The graphic or visual representation formalism that we present here (Paquette, 1996; Paquette,
2002) has been tested for the past 10 years in a vast array of modeling applications and in many
various contexts. It is used by trainers for corporate training, designers or professors use it to
prepare university courses or to propose modeling exercises to their students. It has served to
model processes for the implementation of a computer-supported high school, or to model
instructional methods or research projects processes.
4
Basis for a Graphical Knowledge Representation Language.
It is often said that a picture is worth a thousand words. That is true of sketches, diagrams, and
graphs used in various fields of knowledge. Conceptual maps are widely used in education to
represent and clarify complex relationships between concepts. Flowcharts are graphical
representations of procedural knowledge or algorithms. Decision trees are another form of
representation used in various fields, particularly in decision-making expert systems.
All these representation methods are useful at an informal level, as thinking aids and tools for the
communication of ideas, but they also have their limitations. One is the imprecise meaning of the
links in a model. Another issue is the ambiguity around the type of entities, symbol system that is
used. Objects, actions on objects and statements of properties about them are all mixed-up,
which make graph interpretation a fuzzy and risky business. Another difficulty is to combine
more than one representation in the same model. For example, concepts used in procedural
flowcharts as entry, intermediate or terminal objects could be given a more precise meaning by
developing them in conceptual sub-models of the procedure. The same is true of procedures
present in conceptual models that could be developed as procedural sub-models described by
flowcharts, combined or not with decision trees.
In software engineering, many graphic representation formalisms have been or are used such as
Entity-Relationship models (Chen, 1976), Conceptual Graphs (Sowa, 1964), the Object
Modelling Technique (OMT) (Rumbaugh, Blaha, Premerlani, Eddy& Lorensen, 1991), KADS
(Schreiber, Wielinga & Breuker, 1993) or the Unified Modeling Language (UML) (Booch,
Jacobson & Rumbaugh, 1999). These representation systems have been built for the analysis and
architectural design of complex information systems. The most recent ones require the use of up
to eight different kinds of model and links, which rapidly become hard to follow without
considerable expertise.
Our initial goals were different. We needed a graphic representation system that was both simple
enough to be used by educational specialists, such as teachers, professors and tutors, who are not,
in general, computer scientists, still general and powerful enough to represent the components
and their relationships of computer-based educational environments.
There is a consensus in educational science to distinguish four basic types of knowledge entities
(facts, concepts, procedure and principles), despite some diversity in terminology and
definitions. See for example, the work of Merrill (1994), Romiszowski (1991), Tennyson and
Rash (1988), and West & Farmer and Wolf (1991). This categorization is retained as the basis
for the MOT graphic representation language.
All four types of knowledge are also considered in the framework of schema theory. The concept
of schema is the essential idea behind the shift from behaviourism to cognitivism, the now
dominant theory in psychology and other cognitive sciences, based on the pioneering ideas of
Inhelder and Piaget (1958) as well as Bruner (1973). In the early seventies, Newell and Simon
(1972) developed, on the same basis, a rule-based representation of the human problem solving
procedural activity, while Minski (1975) defined the concept of "frame" as the essential element
to understand perception, and also to reconcile the declarative and procedural views of
knowledge.
Schemas play a central role in knowledge construction and learning (Holoyak, 1991; Anderson
et al 1995). They defined perception as an active, constructive and selective process. They
5
support memorization skills seen as processes to search, retrieve or create appropriate schemas to
store new knowledge. They describe understanding as possible by the comparison of existing
schema with new information. Globally, through all these processes, learning is seen as a schema
transformation enacted by higher order processes, aiming at schema construction and
reconstruction through interaction with the physical, personal or social world, instead of a simple
transfer of information from one individual to another.
The distinction between conceptual and procedural schema has been accepted for a long time in
cognitive science. More recently, a third category called "conditional or strategic schema" has
been proposed (Paris, Lipson, & Wixson, 1983). These schemas have a component that specifies
the context and the conditions to trigger a set of actions or procedures, or to assign values to the
attributes of a concept. These categories map very well on the existing consensus in educational
science
The MOT Visual Modeling Language.
We will now present briefly the syntax and semantic of the MOT visual modeling language,
based on the notion of schema. Here, we could use graphs similar to UML object models to
represent the attributes that describe a schema with different formats according to their type. In
the MOT graphic language (Paquette, 1996, Paquette, 1999, Paquette, 2003), we have improved
the readability and the user-friendliness of graphs by externalizing the internal attributes of a
schema into other objects, with proper links to the original schema or object. For example, the
link between the schemas “Triangle” and the “Rectangle Triangle” is shown explicitly using a
specialization (S) link from the later to the former concept. Links between the “Triangle”
concept and its sides or angles attributes is externalized using a composition (C) link. The links
from an input concept to a procedure and from a procedure to one of its products are both shown
by an input/product (IP) link. The sequencing between actions (procedures) and/or conditions
(principles) in a procedure is represented by a precedence (P) link. Finally, the relation between a
principle and a concept that it constrains, or between a principle and a procedure that it controls,
will be represented by a regulation link (R).
Using these links, this example on triangle concepts becomes the MOT model in figure 1 where
relations between knowledge entities are transparent, mixing the types of entities and links.
Figure 1– A simple MOT model
6
Concepts (or classes of objects), procedures (or classes of actions) and principles (or classes of
statements, properties or rules) are the primitive objects of the MOT graphical language. The
type of the object is represented by geometrical figures as shown on figure 2, where each class or
individual is represented by a name within the figure.
Figure 2 – Types of knowledge units in MOT
These objects are different types of schema whose attributes are all explicitly externalized and
related to the schema using six kinds of typed links constrained by the following grammar rules:
1. All abstract knowledge units (concepts, procedures, principles) can be related by an
instantiation I link to a set of facts representing individuals called respectively examples,
traces and statements.
2. All abstract knowledge units can be specialized or generalized to other abstract knowledge
using specialization S links.
3. All abstract knowledge units can be decomposed, using C links into other entities, generally
of the same type.
4. Procedures and principles can be sequenced together using P links.
5. Concept can be inputs to a procedure using an IP link to the procedure, or products of a
procedure using an IP link from the procedure.
6. Principles can regulate, using an R link, any procedure to provide an “external” control
structure, to constrain a concept or a set of concepts by a relation between them, or to
regulate a set of other principles, for example to decide on conditions of their application.
Figure 3 summarizes these grammar rules of the MOT graphic language in the form of an
abstracted graph where the entities represent types of MOT objects.
Figure 3 – The MOT metamodel
There are various possible semantic interpretations of these graphic symbols.
Concepts can be object classes (country, clothing, vehicles,…), types of documents
(forms, booklets, images,...), tool categories: (text editors, televisions,…), groups of
people (doctors, Europeans,…), or event classes (floods, conferences,…).
Procedures can be generic operations (add numbers, assemble an engine,…), tasks
categories (complete a report, supervise a production,…), activities (take an exam, teach
7
a course,…), instructions (follow a recipe, assemble a device…), or scenarios (of a film,
of a meeting, of a learning module).
Principles can state properties of objects (cars have four wheels), constraints on
procedures (the tasks must be completed within 20 days), cause/effect relationships (if it
rains more than 25 days, the crop will be in jeopardy), laws (any metal sufficiently heated
will stretch out), theories (the laws of the market economy); rules of decision (advising
on an investment), prescriptions (medicinal treatment, instructional design principles),
etc.
The MOT+ Graphic Editor.
With this set of primitive graphic symbols, it has been possible to build graphic models, from
simple to complex representations of structured knowledge. For example, we can build
representations equivalent to conceptual maps, flowcharts (iterative procedures) and decision
trees, and also other types of models useful for educational modeling such as processes, methods
and theories. All these types of models have been used in a number of projects since the first
publication of the MOT editor in 1998, and also in the last five years with its extension to
MOT+. Figure 4 presents examples of the main MISA visual models constructed with the MOT
editor.
Figure 4a – A Knowledge Model
Figure 4a presents an
example
of
a
Knowledge model that
describes part of the
knowledge in the
domain of artificial
intelligence (AI) for an
introductory
Webbased course on that
subject designed with
MISA. Here ovals
represent
AI
processes, rectangles
represent AI concepts
and
hexagons
represent
AI
principles.
8
Figure 4b presents
a example of a
Pedagogical
model represents
a
learning
scenario
model
for one the course
modules
where
learning activities
are represented as
procedures
(ovals)
and
learning resources
as concept/object
(rectangles).
Figure 4c presents
an example of a
Media
model
representing the
structure of a
Web site for the
course. Concepts
represent
Web
pages or page
elements, ovals or
circles represent
hyperlinks,
as
possible actions
or
procedures.
Templates
are
represented
by
principles. Facts
represent concrete
object such as
page
elements
with their actual
texts, pictures or
other resources.
Figure 4b – A Pedagogical Model
Figure 4c – A Media Model
9
Figure 4d – A Delivery Model
Figure 4d presents an
example of a Delivery
model represents the
course
delivery
process where actors
are represented as
control
principles,
acting
on
tasks
represented
as
procedures,
each
having input and
output resources.
This first version of the MOT editor has been extended to the MOT+ editor, a mature editor with
advanced graphic editing capabilities (fonts, color, disposition on a page, etc.). Sub-models can
be embedded at any depth and knowledge objects in each one can be displayed in a multilayer
mode. Models may be filtered in order to display only some types of knowledge objects or links.
Sub-models from one model can be associated to objects in another model called a co-domain,
which is very useful for example to assign knowledge to activities in a pedagogical model.
Graphic objects can be associated to any type of document using the OLE standards such as a
word document, slide presentation, Web page, spreadsheet or database file, which can be
displayed by clicking on the graphic symbol. MOT+ has extensive export facilities to XML,
HTML, Excel and other commonly used formats. In particular, the export to XML command
provides the possibility for graphic models to be processed by software agents respecting for
example the IMS LD or OWL schemas.
REPRESENTING MULTI-ACTOR WORKFLOWS AND LEARNING DESIGNS.
In the two following sections, we address the issues of the standardization of visual modeling
languages, to promote the reusability of educational models and the interoperability between
systems delivering learning environments. With the advent of an educational modeling standard
specification like IMS-LD, we decided to develop a specialization of MOT+ to represent the
IMS-LD concepts. During the eduSource and LORNET (ref) projects, we found that this
specification was closely related to the MISA pedagogical model including some aspects of the
MISA delivery model. This R&D is presented in section 3.1, the extension to a web based
graphical editor is presented in section 3.2.
10
The MOT+LD Special Visual Language.
IMS-LD provides a representation of the components of a learning environment in a
standardized XML schema that can be executed by any compliant eLearning platform. IMS-LD
does not provide a visual language to build a learning environment specification. Initially, these
had to be built using an XML editor or a form-based editor like RELOAD (2005). Also, IMS-LD
is not an instructional design method to build such representations. It needs to be accompanied
by any instructional design method, and MISA is more closely related than many other methods.
Unfortunately, the MOT+ pedagogical models built in MISA are not executable on a variety of
platforms because they are not standardized. In fact, in the projects where we have used MISA,
the specification was translated by hand, into the platform’s activity editor, with some loss of
information.
To address these problems, we first developed a graphic modeling editor for the IMS-LD
specification (level A) and made it available as specialized editor in the MOT+ software. Many
examples of learning designs have been produced by different groups using this editor. They can
be found at the IDLD portal (2006). Figure 5 shows part of a simple example of a Unit of
Learning (UoL) on solar astronomy presented recently at a workshop (Paquette & Léonard,
2006).
Figure 5 – An example of a MOT+LD learning design
It shows an act and its activity structure containing various learning and support activities, all
represented as MOT procedures (ovals). Since method, plays and acts, as the IMS-LD metaphor
applies as concepts for an instructional structure, are also represented as procedures in other parts
of the model. Each procedure type is indicated by little label at the right lower corner of the ovals
representing the procedures.
Similarly, roles are represented by different kinds of MOT principles (hexagons). Environments,
learning objects, services and outcomes are represented by different kinds of MOT concepts
(rectangles). Standard MOT links are used between these objects. C (is composed of), P
11
(precede), R (regulate or govern) and I/P (input / product) links are sufficient to cover all the
components of a standard IMS-LD level A learning design.
The MOT+LD editor is presented with some detail in (Paquette, Léonard, Lundgren-Cayrol,
Mihaila & Gareau, 2006). It enables a designer to build graphically a compliant IMS-LD model.
Afterwards, the graph is automatically validated and exported as an instance of the IMS-LD
XML schema. This XML file can be read in form-based IMS-LD editors such as RELOAD
(2005), if level B conditions and or level C notifications need to be specified. The XML can then
be run by IMS-LD compliant players or platforms to deliver on-line learning sessions to their
users.
Paquette & Marino (2005) briefly discuss the strengths and weaknesses of the IMS-LD
educational modeling specification. One weakness is the absence of knowledge representation,
which is central to learning and knowledge management that we seek to support by the TELOSiii
system. We have proposed to improve that by the semantic annotation of the activities, resources
and roles included in a learning design. A semantic annotation is a mapping from a subject
matter ontology to the learning design that associates knowledge elements to the components of
the design. This aspect will be developed in the following sections.
Extending the MOT+LD Editor.
Another aspect of IMS-LD we need to improve is the control structure of the workflow, that is
actually covered by level B and C specifications, where properties and conditions can be
included in the design to alter the flow of activities, notify an actor or present a resource
depending on previous actions or results stored in a user and group file or model. This aspect
may not be that important in open learning environments where a total or large degree of liberty
is left to the learner and facilitators, but for a business workflow in an organization, or to
aggregate software components into larger resources, it is an important dimension.
To address that and provide a basis to build a function editor for the TELOS system, conceptual
work on function maps has been defined as a central piece of the TELOS architecture (Rosca,
2005; Paquette, Rosca, Mihaila & Masmoudi, 2006). Moreover, a comparative analysis has been
made between business workflows, IMS-LD learning designs and function maps (Marino et al,
2006), leading to the identification of 21 control situations for workflows encountered in
software engineering literature (Correal & Marino, 2006). It was found that IMS-LD covers only
some of these control situations, but probably the most useful ones for pedagogical design.
Based on this work and the actual MOT+LD editor, we are in the process of designing a new
visual editor. The Function Editor aims both to generalize IMS-LD and to capture the main
aspects of business workflows. The graphs produced by this editor will be used as executable
interfaces for concrete actors to enact the activities and use the resources during delivery. It will
also serve to orchestrate actors, activities and other resources, a fundamental principle built in to
the TELOS system. A specialization of the function editor is being defined to cover all three
levels of the IMS-LD specification.
The Function editor uses four kinds of MOT objects with subtypes taken from the TELOS
technical ontology (Magnan & Paquette, 2006). These are shown on figure 6. Concept symbols
represent all kinds of resources: documents, tools, semantic resources, environments, resourceactors, resource-activities and datatypes. Procedure symbols represent activities, including
function models or commonly used operation templates to be embedded in other activities.
12
Finally, principles are used both to represent different types of actors (as control agents) and
control conditions. These two kinds of control entities are represented here by different symbols.
The actor’s symbols are active agents representing users, groups, roles or software agents that
enact the activities using and producing resources as planned by the function model. Conditions
are control element inserted within the basic flow to decide on the following activities that can
be activated.
Figure 6 – Function Editor Symbols
Figure 7 – A simple function model
In figure 7, we see a combination of some of these symbols where a coordinator writes the plan
of a document in activity-0. After that the figure shows a general split condition after activity-0.
After that, activities 1, 2 and 3 are executed in parallel, controlled by the properties of the split
condition object. Later on, the flow of activities merges through the merge condition object
before activity M+1 takes control. This activity will wait for some or all the incoming flows to
be activated before it is executed, again based on the properties of the merge condition object.
The basic control flow is shown by P links and it is altered by R links. The data flow is shown by
IP links.
Figure shows another kind of
condition that alters the flow of
execution. In activity-2, if the timeevent condition is met, the flow of
control will change. Depending on
the type of the condition, the
activity-4 will be shown or hidden.
Activity-3 is still available. If
activity-4 is shown and completed,
then activity-5 can be performed.
Properties of the event condition
symbol will provide the details on
the condition and action parts of the
control principle to provide the
execution engine with a clear
formal definition of the processing
to take place.
Figure 8– Event-based control
13
In the Function Editor, we see a combination of a control flow and a data flow. The control flow
is modeled using the MOT basic P and R links. P links indicates the basic sequence or flow of
activities. R linked conditions identify which activities an event will trigger, thus altering the
basic flow.
IP links from MOT serve to model the data flow, either from resources to activities where they
are consulted, used or processed , or from activities to the resources they help produce. This is
why we need to distinguish between actors as active control entities and resource-actors that will
serve as data providers or be products of an activity (e.g. a new person of software agent added
in a system). A similar distinction is made for resource-activities that can be seen as resources to
be transformed, for example by other activities creating or modifying their description.
C links from MOT may also be used to show the composition of an entity into other entities. A
new unification, U link, is also necessary to guide the execution engine, when components are
aggregated.
In TELOS, the function editor will enable engineers to combine resources into larger aggregates,
technologist to built platform workflows for designers of learning or knowledge management
environments, designers to build courses, work flows or learning /teaching scenarios.
MOT+OWL: A STANDARDIZED ONTOLOGY EDITOR.
In section 1, we identified the pedagogical and the knowledge models as the most important
ones. We now proceed with a second standardization task, that of the knowledge model. Any
type of knowledge representation, including text-based narratives or informal graphic models,
can be used to describe a domain of study. At the initial stage of design, the informal nature of an
ontology representation is useful. The user’s mind must be free to choose any representation that
seems best suited for the educational project to be considered. Still, this very freedom does not
facilitate the software processing of the representation.
Semi-formal modeling languages like MOT go part of the way in that direction. Unlike informal
graphs built with any graphic editor, such as PowerPoint, the MOT graphic syntax is structured
and has a general unambiguous semantic. Using the MOT editor, models can be exported in
many formats, including a native XML schema. Using this schema, software agents can perform
different kinds of processing. Still, some ambiguity remains. In instructional engineering
applications, we had to constrain the MOT graphic language even more to enable the delivery of
learning scenarios in a digitized platform like Explor@-2 (Paquette, 2001). Even then, part of the
transfer of the design to the delivery platform had to be done manually, to prevent enforcing
unnatural graphic representations on the users.
The Ontology Web Language.
To deliver computer-based learning environments, after a phase where informal graphic design
has cleared up ideas, we need to move from informal or semi-formal graphs to formal
computable graphic representations. Knowledge in a subject domain can be represented in many
ways: taxonomies, thesauri, topic maps, conceptual graphs and ontologies.
We have selected to start with OWL-DL ontologies (see W3C, 2004) for a number or reasons. It
is one of the three ontology Web languages that are part of the growing stack of World Wide
Web consortium recommendations related to the Semantic Web. Of these three languages,
OWL-DL has a wide expressivity and its foundation in descriptive logic guarantees its
14
computational completeness and decidability. Descriptive Logic (Baader, Calvanese, Nardi,
Patel-Schneider, 2003), is an important knowledge representation formalism unifying and giving
a logical basis to the well known traditions of frame-based systems, semantic networks, objectoriented representations, semantic data models, and formal specification systems. It thus
provides an interesting framework to represent knowledge for which a growing number of
processing agents are built throughout the world.
OWL-DL provides a precise XML schema but no graphic representation per se. Some ontology
editors like PROTÉGÉ (2006), provide some graphical views of the ontology, but the
construction of an ontology is essentially form-based. Our goal was to provide a complete formal
graphic representation of the OWL-DL that could combine the virtues of interactive construction
with the computational capabilities of a formal graphic representation.
The MOT+OWL Visual Language.
In the context of the MOT representation system, ontologies, in particular OWL-DL constructs,
correspond to a category of models called theories. Ontologies can thus theoretically be modeled
graphically using the MOT syntax. While doing this, we found out that while the MOT primitive
objects and links were sufficient to represent ontologies expressed in OWL-DL, the graphs
would become cumbersome unless new symbols were added. We have thus specialized the MOT
language and graphic editor by adding sub-types for concepts, principles and facts and by adding
new links.
Table 2 gives a few examples of the MOT+OWL graphic elements with their interpretation in
descriptive logic and their correspondence to standard OWL-DL XML schema fragments. See
(Paquette & Rogozan, 2006) for a complete description of the MOT+OWL graphic language.
Three types of MOT entities are sufficient to represent OWL-DL models. Concepts represent
classes, principles represent properties and facts represent individuals. On these graphic entities,
icons are added corresponding to axioms or principles stating a property of the class. We also
added new special links to express things like equivalent “equi” or disjoint “disj” classes stating
properties of two classes or two properties.
In the standard MOT syntax, these icons or special links would be expressed by principles with
“R” links to classes or properties. For example, in the second and the two last examples of Table
2, the following standard graphs (figure 9) are equivalent, with the same precise OWL-DL
interpretation as XML schema components. These would of course make the graphs more
difficult for human interpretation.
Are
équivalent
R
Class1
R
Class2
Is
transitive
R
Property 1
Are
inverse
R
Property 1
R
Property 2
Figure 9 –MOT standard equivalents
15
Table 2 - OWL-DL equivalents
MOT+OWL graphic
symbol
Description logic
statements
Class intersection
∀x: Class3(x) ↔
Class1(x) ∧ Class2(x)
Equivalent classes
Class1
Equi
Class2
Class1
Disj
Class2
∀x: Class1(x) ↔ Class2(x)
Disjoint classes
∀x: Class1(x) ↔
¬Class2(x)
OWL-DL XML-Schema segment
owl:Class>
<owl:intersectionOf rdf:parseType="Collection">
List of class descriptions
</owl:intersectionOf>
</owl:Class>
<owl:Class rdf:about="#name_class1">
<equivalentClass rdf:resource="#name_class2"/>
</owl:Class>
<owl:Class rdf:about="#name_classe1">
<owl:disjointWith
rdf:resource="#name_classe2"/>
</owl:Class>
∀x: Class(x) ↔ (x = Ind 1)
∨...∨ (x= Ind N)
<owl:Class>
<owl:oneOf rdf:parseType="Collection">
<owl:Thing rdf:about="#name_individual1"/>
<owl:Thing rdf:about="#name_individual2"/>
…
<owl:Thing rdf:about="#nom_individualN"/>
</owl:oneOf>
</owl:Class>
Functional
property
<owl:FunctionalProperty
rdf:about="#name_property" />
Extension of a
class
Prop(x,y) ∧
Prop(x,z)) → y=z
∀x,∀y,∀z:
Transitive
property
<owl:TransitiveProperty
rdf:about="#name_property" />
∀x,∀y,∀z: Prop1(x,y) ∧
Prop1(y,z) → Prop1(x,z)
Inverse properties
∀x,∀y: Prop1(x,y) ↔
Prop2(y,x)
<owl:ObjectProperty rdf:ID="name_Property1">
<owl:inverseOf
rdf:resource="#name_property2"/>
</owl:ObjectProperty>
Using a limited set of graphic symbols, we can formally describe any semi-formal MOT model
that is amenable to a representation in descriptive logic. This is obviously the case for most
conceptual models, laws and theory models. However, this is less evident in the case of
procedural models, sometimes called task ontologies. Procedural and process/methods models
are important for our purpose because learning environments are built around multi-actor
processes.
Figure presents a MOT+OWL visual graph that translates the conceptual structure of a learning
design presented in the IMS-LD information model (2003). In the figure, “C” properties (green
hexagons) are an abbreviation for “is-composed-of” which has the same meaning as the C link in
standard MOT models, or the aggregation link in UML models.
16
This example illustrates the fact that functional relations between components of multi-actor
processes such as a learning design can be represented by ontologies. Such ontologies have been
used to test, for example, the conformance of particular learning designs to the IMS-LD XML
schema (Amorim, Lama, & Sanchez, 2006), and to execute them in the context of an ontologydriven system.
Figure 10 – A simple task ontology for multi-actor scenariosiv
ASSOCIATING KNOWLEDGE & COMPETENCIES TO LEARNING DESIGNS.
We have pointed out earlier the importance of associating knowledge and competencies
(semantic annotation) to the components of a learning design. This is a key element of the MISA
method. Actually, in IMS-LD, the only way to describe the knowledge needed to achieve the
activities or that is present in the resources is to assign optional educational objectives and
prerequisites, to the unit of learning as a whole and/or to all or some of the learning activities,
but can not be added to express the level of competency for a support activity carried out by a
teacher or tutor. Objectives and prerequisites correspond to entry and target competencies as
used in the MISA method. They are essentially unstructured pieces of text composed according
to the IMS RDCEO specification (IMS 2002).
17
Unstructured texts are difficult to compare. Consistency checking between different levels of the
LD structure cannot be supported computationally. Even at the same level of a learning design,
for example within an act, no relations exist between the content of learning activities and of the
input or outcome resources, and from these to the actors’ competencies. In fact, in IMS-LD the
knowledge represented in learning resources is not described at all, and the actor’s knowledge
and competencies are only indirectly defined by their participation in learning units or activities,
if, and only if, educational objectives have been associated to the activities.
What we need first is a qualitative structural representation of knowledge and competencies
associated to activities, resources and roles. This can be done using domain ontologies. As a first
step, the MOT+ editor allows to show side by side a learning design, as well as associating it to a
co-model, using the MOT+LD editor, and a domain knowledge ontology using the MOT+OWL
editor. An example is shown in Figure . The left hand window is the learning design presented
earlier in Figure . The right hand window presents part of domain ontology of the solar system
(that was built before Pluto was declared a quasi-planet).
A semantic annotation is simply a mapping from the domain ontology to the learning design that
associates knowledge elements (classes, properties and individuals of the ontology) to
components of the learning design.
Figure 11 – An example of ontology annotation of a learning design
In Figure , we see that data on the orbital period of planets in the solar system has been
associated to a learning object in the design, which is a PowerPoint presenting this data to team
A. This resource is an input to learning activity 2.1.A, but it is not the only input to this activity.
There is also another resource (clues A) that gives additional information to team A, plus the
chat between team members that will bring other clues to each participant. As a result, the submodel of the ontology associated to activity 2.1A would logically correspond to the union of the
sub-models of all input resources to the activity.
Finally, the figure shows that most of the ontology model should be the subject of the discussion,
since there is another team, team B that has more information to bring to the discussion using
18
also information from input resources and in a team B chat. The larger sub-model is thus
associated to the 2.0 activity structure.
This example shows how semantic annotation can help guide the construction of learning
designs and to evaluate their coherence. By associating the right amount of knowledge to the
different resources and activities, a designer can build a coherent design that will trigger
collaboration between learners, or help a trainer decide on its intervention, or guide the actions of
an intelligent tutoring system, and, in general support the evolution of the learners’
competencies.
DESIRABLE PROPERTIES IN A VISUAL EDUCATIONAL MODELING LANGUAGE.
This chapter concludes with a discussion of the most important features and characteristics of a
visual educational modeling language, which we think are the most useful and beneficial to the
user.
Visual. The benefits of graphical cognitive modelling have been eloquently summarized by
Ausubel (1968), Dansereau (1978), Novak (1993) and Jonassen, Beissner & Yacci (1993).
Graphs illustrate relationships among components of complex phenomena. They uncover the
complexity of actors’ interactions and makes the most important parts stand out. They facilitate
the communication about the reality studied. They favour the global comprehension of the
phenomena under study. They help grasp the structure of related ideas by minimizing the use of
ambiguous natural language texts. As an example, entity-relation graphs reduce ambiguity
compared to a natural language description, but some remain on the interpretation of the terms
written on the links or on the nodes. Ambiguity can be reduced further by the use of standardized
typed objects and typed links.
User-friendliness. Not all graphic modeling languages are user-friendly. A good counterexample is UML. The large number of models and symbols require considerable expertise and a
steep learning time for the interpretation and for the construction of models. Furthermore, each
type of model captures a different viewpoint of the information and it is impossible to mix them
in the same graph to provide a global view of a subject domain. The representational system
must be easy to use without technical or scientific mastery after a short period of initiation.
Dansereau and Holley [39], have studied experimentally the use of different sets of graphic
symbols by learners. Their results show that typed links are preferred by the majority of learner,
as long as there are not two few nor two many links and they express sufficiently different
meanings.
Generality. Generality means that the representation language should have the capacity to
represent, with a relatively small number of object and link categories, all knowledge in very
different subject domains, at various levels of granularity and precision. It should enable, to
represent simple models such as a multiplication table, up to complex models such as multi-actor
workflows, rule-based knowledge systems, methods and theories. It should also embed
equivalent representations to commonly used graphs such as conceptual maps, semantic
networks, flowcharts, decision trees or cause/effect diagrams. .
Formalizable. The graphic language should be upward compatible from informal graphs, up to
semi-formal and totally unambiguous formal models. At the informal level, an integrated
representation framework facilitates the organization of thought and communication between
humans about the knowledge which is exchanged, all along the evolution of the graphic
19
representation model. Here the process is more important than the result. On the other end, the
graphic language makes it possible to use more constrained elements to produce totally
unambiguous descriptions that can be exported to set of symbols, such as an XML file, to be
processed by computer agents. Here the model is more important than the process.
Declarative. Graphic language can be procedural or declarative. Procedural graphic languages
have been built in the past; essentially extending flowcharts to promote graphical programming
that would produce code directly. Our proposal is to use, as much as possible, a declarative
graphic language, for a number of reasons. Firstly, it is easier for a person to declare the
components of his/her knowledge than to describe the way it should be processed. In expert
systems for example, the executive instructions are not wired-in the program, but externalized
and made visible in a knowledge base on which a general inference engine proceeds. Secondly,
the same model can be used for many different applications, not necessarily the one for which
the processing has been planned in a procedural program. This is done by querying the model
using an inference engine, in a Prolog-like manner. Thirdly, the processing knowledge itself can
be given declaratively, so that higher order meta-knowledge, also can be singled-out. This idea is
similar to structural analysis as proposed by Scandura (1973) and it is exactly the way we should
see the relation between generic skills and domain knowledge in a competency, as metaknowledge given declaratively, applied to domain knowledge, for example, rules for diagnosing
a component-based system applied to different models describing a car, a software or a learning
environment.
Standardized. Standardization is an important property to enlarge knowledge communication
and use between persons or software agents. At the informal level, each model constructed by a
person must be interpretable by another person. At the formal level, the communication
capabilities extend to software agents. The move towards graphic versions of standards like
IMS-LD for learning designs and OWL for ontologies adds wider communication capabilities
between researchers and educators while at the same time adding formal non-ambiguous
interpretation for machine processing.
Computability. Computability is a step beyond standardization. Not only can the graphic model
receive a non-ambiguous formal representation that can be processed by computer agents, but
this formal representation is complete (all conclusions are guaranteed to be computable) and
decidable (all computations will finish in finite time). These considerations have motivated the
construction of the MOT+OWL graphic language that is equivalent to the OWL-DL XML
schema based on descriptive logic. OWL-DL ontologies are declarative, and standardized by the
W3C.
CONCLUSION
This chapter has presented a 10-year effort to provide an educational visual language for
applications that can span form informal support to idea generation, up to structured semi-formal
graphs based on typed objects and links, and finally to graphic design on the formal conceptual
and specification levels (MOT+LD, MOT+OWL).
In Botturi et al (2006), the reader can find a classification of other visual languages, some of
them being presented in other chapters of this Handbook. According to this classification, MOT+
has the same properties as those of UML. It qualifies as a visual, layered, formal, conceptual and
specification elaboration language, with multiple perspectives.
20
This correspond to our initial goal of building a virtual language that is both user-friendly for
designers (compared to UML) and still general and powerful enough to enable the design of the
main components of a learning system, according to standard specifications. With the
development of the new function editor based on MOT+ concepts, we can now go a step further
and provide a visual scenario programming language that can be executed by an ontology-based
engine to deliver usable learning environments to its users.
REFERENCES
Amorim R., Lama, M. & Sanchez, E. (2006). Using Ontologies to model and execute IMS Learning
Design Documents.. The 6th IEE International Conference on Advanced Learning Technologies
(ICALT-06), (pp.115-116). Kerkrade, The Netherlands.
Andersson, J.R. Corbett, A.T., Koeudinger, K.R. & Pelletier, R. (1995). Cognitive Tutors: Lessons
learned. The Journal of Learning Sciences, 4 (2), 167-207.
Ausubel, D. P. (1968). Educational Psychology; A cognitive view. New York, Rhinehart & Winston.
Baader, F., Calvanese, D., McGuinness, D., Nardi, D., Patel-Schneider, P. (Ed.). (2003). The Description
Logic Handbook. Cambridge University Press.
Booch.G., Jacobson, J. & Rumbaugh, I. (1999). The Unified Modeling Language User Guide. USA:
Addison-Wesley.
Botturi L., Derntl M., Boot, E. and Figl K. (2006) A Classification Framework for Educational Modeling
Languages in Instructional Design. The 6th IEE International Conference on Advanced Learning
Technologies (ICALT-06), (pp.1216-1220) Kerkrade, The Netherlands.
Bruner, J.S. (1973).Beyond the information given. New York: Norton.
Chen, P.P.S (1976).The Entity-Relationship model - toward a unified view of data. ACM Transactions on
Database Systems I, 1 (1), 9-36.
Correal. D., & Marino O. (2006). Software Requirements Specification Document for General Purpose
Function’s Editor (Technical Report V0.4), Montréal, Canada: Télé-université LICEF Research
Centre, Montreal.
Dalziel, J.R. (2005).LAMS. Learning Activity Management System 2.0. Retrieved January 22, 2007
from http://wiki.oamsfoundation.org/display/lams/Home
Dansereau D.F. & Holley, C.D. (1982).Development and evaluation of a text mapping strategy. In A.
Flammer & W Kintsch (eds.), Discourse Processing. The Netherlands: North Holland.
Dansereau D.F. (1978).The development of a learning strategies curriculum. In H. F. O’Neil Jr.,
(ED.).Learning strategies. New York: Academic Press. Davies,
Holoyak, K.J. (1991). Symbolic connectionism: Toward third generation-theories of expertise. In K.A.
Ericsson & J. Smith, Toward a general theory of expertise: Prospects and Limits. New York:
Cambridge University Press.
IDLD (2006) Implementation and Deployment of the Learning Design Specification Portal.
http://www.idld.org
IMS-LD (2003). Information Model, Best Practice and Implementation Guide, Binding document,
Schemas. IMS Learning Design. Retrieved October 3, 2003, from
http://www.imsglobal.org/learningdesign/index.cfm
Inhelder, B. & Piaget, J. (1958).The growth of logical thinking from childhood to adolescence. New
21
York: Basic Books
Jonassen D.H., Beissner K., & Yacci M. (1993).Structural Knowledge – Techniques for Representing,
Conveying and Acquiring Structural Knowledge. New Jersey: Laurence Earlbaum Associates
Magnan, F., & Paquette, G. (2006). TELOS: An ontology driven elearning OS. In Weibelzahl, S.,
Cristea, A., (Eds.), Workshops held at the Fourth International Conference on Adaptative
Hypermedia and Adaptative Web-Based Systems 2006. (pp. 131–139). Dublin, Ireland: National
College of Ireland.
Marino, O., Casallas R., Villalobos J., Correal D. & Contamines, J. (2006).Bridging the Gap between elearning Modeling and Delivery through the Transformation of Learnflows into Workflows , in S.
Pierre (Ed).E-Learning Networked Environments and Architectures: a Knowledge Processing
Perspective, Springer-Verlag.
Merrill, M.D (1994). Principles of Instructional Design. Educational Technology Publications, New
Jersey: Englewood Cliffs.
Minski, M. (1975). A framework for representing knowledge In P. H. Winston (ED.), The psychology of
computer vision. New York: McGraw-Hill
Newell, A. & Simon, H. (1972). Human problem solving. NF: Englewood Cliffs.
Novak, J. D. (1993). How do we learn our lesson? Taking students through the process. The Science
Teacher, 60(3), 50-55.
OMG (2006). Business Process Modeling Notation (BPMM). Retrieved on July 24th, 2006 from
http://www.bpmn.org/
Paquette, G. (2004 ). Instructional Engineering for Network-Based Learning. USA: Pfeiffer/Wiley
Publishing Co.
Paquette, G. & Léonard, M. (2006). The Educational Modeling of a Collaborative Game using
MOT+LD. Proceedings oFthe the 6th IEE International Conference on Advanced Learning
Technologies, (pp.115-116), Kerkrade, The Netherlands.
Paquette, G. (1996).La modélisation par objets typés: une méthode de représentation pour les systèmes
d’apprentissage et d’aide à la tâche. Sciences et techniques éducatives, (4),9-42
Paquette, G. (2001). Designing Virtual Learning Centers. In H. Adelsberger, B. Collis, J. Pawlowski
(Eds).Handbook on Information Technologies for Education & Training (pp. 249-272). The
Netherlands: Springer-Verlag.
Paquette, G. (2002). TeleLearning Systems Engineering – Towards a new ISD model. Journal of
Structural Learning 14, 1-35.
Paquette, G. & Marino, O. (2005). Learning Objects, Collaborative Learning Designs and Knowledge
Representation, Technology, Instruction, Cognition and Learning, 3, 85-108.
Paquette, G. & Rogozan, D. (2006). Primitives de représentation OWL-DL - Correspondance avec le
langage graphique MOT+OWL et le langage des prédicats du premier ordre. TELOS
documentation. Montreal, Québec: LICEF Research Center.
Paquette G., Marino, O., De la Teja, I., Lundgren-Cayrol, K., Léonard, M. & Contamines (2005).
Implementation and Deployment of the IMS Learning Design Specification, Canadian Journal of
Learning Technologies,31(2) (CJLT), http://www.cjlt.ca/
Paquette, G. (2002). Modélisation des connaissances et des compétences, un langage graphique pour
concevoir et apprendre. Québec, Canada : Presses de l’Université du Québec.
22
Paquette, G. (1999). Meta-knowledge Representation for Learning Scenarios Engineering, in S. Lajoie et
M. Vivet (Eds), Proceedings of AI-Ed’99 in AI and Education - Open learning environments, IOS.
Paquette, G. & Rosca. I. (2004). An Ontology-based Referencing of Actors, Operations and Resources in
eLearning Systems. SW-EL/2004 Workshop. The Netherlands: Eindhoven.
Paquette G., Léonard M., Lundgren-Cayrol K., Mihaila S. & Gareau D. (2006). Learning Design based
on Graphical Knowledge-Modeling. Journal of Educational technology and Society ET&S ,
Special issue on Learning Design, January 2006. (Also published in the Proceedings of the
UNFOLD-PROLEARN Joint Workshop, Valkenburg, The Netherlands, September 2005 on
Current Research on IMS Learning Design.)
Paquette, G., De la Teja, I., Léonard, M., Lundgren-Cayrol, K., & Marino, O. (2005). How to use an
Instructional Engineering Method and a Modelling Tool (pp. 161-184), in R. Koper & C. Tattersall
(Eds.). Learning Design - A Handbook on Modeling and Delivering Networked Education and
Training, Springer-Verlag,
Paquette, G., Crevier F. & Aubin, C. (1994).ID Knowledge in a Course Design Workbench. Educational
Technology, (34) 9, 50-57.
Paquette, G., Rosca. I, Mihaila S. & Masmoudi A. (2006 in press).TELOS, a Service-Oriented
Framework to Support Learning and Knowledge Management, in S. Pierre (Ed).E-Learning
Networked Environments and Architectures: a Knowledge Processing Perspective. The
Netherlands: Springer-Verlag
Paquette,G., Marino, O., De la Teja, I., Léonard, M., Lundgren-Cayrol, K., (2005). Delivery of
Learning Design: the Explor@ System’s Case (pp. 311-326). In R. Koper & C. Tattersall
(Eds.). Learning Design – A Handbook on Modelling and Delivering Networked
Education and Training. The Netherlands: Springer Verlag.
Paris S., Lipson M.Y., & Wixson K.K. (1983). Becoming a strategic reader. Contemporary Educational
Psychology, 8, 293-311.
PROTÉGÉ (2006).Protégé Homepage: Description and download available. Retrieved July 24, 2006
from http://protege.stanford.edu/
RELOAD (2005).RELOAD Homepage: Editor and Player. Retrieved July 24, 2006 from
http://www.reload.ac.uk/.
Romiszowski, A. J. (1981). Designing Instructional Systems. New York, USA: Kogan Page
London/Nichols Publishing.
Rosca, I. (2005). TELOS Conceptual Architecture. (LORNET Technical Report: 0.5.).Canada: LICEF
Research Centre, Télé-université.
Rumbaugh J., Blaha M., Premerlani W., Eddy F., & Lorensen W. (1991). Object-Oriented Modelling and
Design. USA: Prentice Hall.
Scandura, J.M. (1973).Structural Learning I: Theory and research. London/New York: Gordon &
Breach science Publishers.
Schreiber G., Wielinga B., & Breuker J. (1993). KADS – A Principled Approach to Knowledge-based
System Development. San Diego, USA: Academic Press.
Sowa, J.F. (1984). Conceptual Structures, Information Processing in Mind and Machine. USA: AddisonWesley Publishing Co.
Tennyson, R. & Rasch, M. (1988).Linking cognitive learning theory to instructional prescriptions.
Instructional Science, 17, 369-385.
23
W3C (2004).OWL Overview Document. Retrieved February 10, 2004 from
http://www.w3.org/TR/2004/REC-owl-features-20040210/ .
West, C. K., Farmer, J. A., & Wolff, P. M. (1991). Instructional design: Implications from Cognitive
Science. Englewood Cliffs, NJ: Prentice Hall.
i
MISA, “Méthode d’ingénierie des systèmes d’apprentissage” is a French acronym meaning « Method for
Instructional Systems Engineering »
ii
ADISA, “Atelier distribué d’ingénierie des systèmes d’apprentissage” is a French acronym meaning « Distributed
Workbench for Learning Systems Engineering ».
iii
TELOS (TEleLearning Operating System) is a new system built within the LORNET project (www.lornet.org) to
enable engineer and technologists to assemble eLearning and knowledge management platforms and environments.
iv
On figure 10, principles with 1 express OWL cardinality axioms here meaning “at least one”.
24