CROSS-DOMAIN MAPPING: QUALITY ASSURANCE AND
E-LEARNING PROVISION
Hilary Dexter, Jim Petch
The e-Learning Research Centre, University of Manchester186 Waterloo Place, Manchester, United Kingdom
Keywords: e-learning provision, quality assurance,
checklist, process model, knowledge domain
Abstract: In order to ensure that a valid and robust model of e-learning provision is developed it has to be based on a
thorough understanding of the e-learning provision domain. The fullest and most detailed articulations of
the e-learning development process are found in quality checklists for e-learning development. The problem
this paper addresses is that posed by the situation of having knowledge used for modeling in one domain
represented by artifacts in another. Using a number of checklist sources, a composite list was developed for
some aspects of the e-learning development process. The checks address the activities and their artifacts that
should be monitored, and what the outcomes of the checks should be in terms of what actions should be
taken and what changes made if the results do not meet quality criteria. A small worked example of this
cross-domain mapping process is given.
1 CONTEXT
The stimulus for this study is the need to ensure
quality service provision for e-learning in higher
education, viz. the processes of planning, design,
development and delivery of e-learning courses.
Underlying the study is an approach to service
provision based on enterprise models. The e-learning
provision model is seen as part of an enterprise
model that includes business processes and
enterprise information model as well as the
provision of e-learning by partner institutions
(Figure 1).
The main premise of this study is that in order to
ens
ure that a valid and robust model of e-learning
provision is developed it has to be based on a
thorough understanding of the e-learning provision
domain. There are two challenges here. One is that
there is no thorough articulation of the e-learning
provision domain that is in any way comprehensive.
The second is that there are very few published
accounts of quality on which to base a model.
Almost all Higher Education (HE) provision is in
situations that are not adequately documented and
the few available commercial sources are
understandably thin.
In fact the fullest and most detailed articulations
of t
he e-learning development process are found in
quality checklists for e-learning development. It
seems that a number of organizations and
individuals have used this means of expression as a
way of capturing and organizing knowledge about
the domain (Scienter-MENON 2004,WCET 2000).
Studies of some of the most widely used and
wel
l known checklists (Hirumi 2003, Franklin,
Petch, Armstrong and Oliver 2004) show clearly that
the scope of these checklists differs substantially and
that the nature of the checks themselves is not
consistent. However it is possible to rationalize the
available checklists (Petch 2003, 2004) so that a
consistent and comprehensive description of the e-
learning development and delivery process is
achieved.
Recognizing that the development of checklists
i
s an ongoing process, a set of published lists was
used to develop a consolidated and harmonized list
that could be used as the basis for developing an e-
Learning Provision model. In this study the list does
not cover the complete e-learning development cycle
(Wilcox, Petch and Dexter, 2004) but is sufficient
for the purpose of exploring the cross-domain
mapping issues.
2 PROBLEM
The problem this paper addresses is that posed by
having knowledge used for modeling in one domain
represented by artifacts in another. It is the problem
199
Dexter H. and Petch J. (2005).
CROSS-DOMAIN MAPPING: QUALITY ASSURANCE AND E-LEARNING PROVISION.
In Proceedings of the Seventh Inter national Conference on Enterprise Information Systems, pages 199-205
Copyright
c
SciTePress
of cross-domain mapping. It is necessitated by the
fact that the best available knowledge of the e-
leaning process is in the Quality Assurance (QA)
domain but the model needed to be developed is in
the provision domain. In fact, in practice there are
iterations of interaction between these domains, and
in some organizations a reasonable expectation that
they have been planned together, so that we may
expect some alignment between them. However,
there remains the problem for the modeler of
developing a satisfactory meta-model and a valid
model in one domain from knowledge based on the
meta-model and model in another. There is no
intention here of modeling the QA domain. The
checklists are taken as given.
The problem of cross-domain mapping is put
forward as a general one for domain modeling. It is
suggested that the situation of asymmetric
positioning of knowledge and model is
commonplace. Indeed a cross-domain mapping
approach may be a useful element of a modeling
strategy in general.
3 APPROACH
A modeling approach has been adopted to tackle the
transfer of knowledge between the two domains of
interest. A modeling framework has been set up to
provide an environment in which it will be possible
to progress in iterations of modeling activity towards
a complete and precise expression of all the people,
processes and technology involved in the provision
of e-learning services. The modeling framework
includes an evolving well-defined vocabulary of
modeling elements expressed in the Unified
Modeling Language (UML).
Figure 1: Modeling the Quality Assurance (QA) and the eLearning Provision domains
A domain model describes the elements that can
exist in the domain, their interrelationships and their
types. Both the static and dynamic aspects of that
domain need to be represented in the model, that is
both the data and information entities and the
business processes. The UML domain model
comprising Classes, Relationships, Use Cases,
Activities and States is equivalent to a formal
ontology for that domain, taking the definition of an
ontology as being “an explicit formal specification
of how to represent the objects, concepts and other
entities that are assumed to exist in some area of
interest and the relationships that hold among them”
(International DOI, 2005). UML may be extended
by stereotypes and tagged values if required to
define precisely concepts in the domain (Fuentes and
Vallecillo, 2004) thus negating the need for a
separate and different ontology language. The
extended UML elements are packaged together into
what is termed a UML profile. In this way an e-
learning profile for UML can be constructed and
added to as more information about the domain is
gathered. This profile may then be applied to any
modeling effort concerned with e-learning provision.
The domain model for e-learning provision
being developed in this research program employs
Class Diagrams, Activity Diagrams and Use Case
ICEIS 2005 - SPECIAL SESSION ON EFFICACY OF E-LEARNING SYSTEMS
200
Table 1: Knowledge Areas Covered by Checklists
ORGANIZATION ACE:
American Council on
Education 1997
AFT:
American Federation of
Teachers 2000
ADEC:
American Distance
Education Council 2004
INSTITUTIONAL
GUIDELINES
o Organizational
Commitment.
o Encourage
experimentation
o Administrative &
organizational
commitment.
PROGRAM DESIGN
AND CURRICULUM
GUIDELINES
o Learning Outcomes
o Technology
o Class size
o Student assessment
o Full programs
o Evaluation of
Coursework
o Technological and
human infrastructure.
COURSE DESIGN
AND
PEDAGOGICAL
GUIDELINES
o Outcomes
o Content
o Expectations
o Interactions
o Assessment
o Complement
Elements
o Technology
o Activities and
assessments
o Potentials of medium
o Personal interaction
o Courses materials
o Outcomes and objectives
o Learner engagement
o Media Use
o Learning environments
o Learning experiences
o Social mission
STUDENT AND
ACADEMIC
SUPPORT
GUIDELINES
o Learner Support o Student requirements
o Advisement
o Research opportunities
o Learner Support
FACULTY
SUPPORT
GUIDELINES
o Academic control
o Faculty Preparation
o Materials Control
Diagrams as a useful subset of the range of tools
available in the UML. The Activity Diagrams
include the flow of artifacts in the domain, and their
state at any stage in the process may be included in
the model. In this way the lifecycles of significant
artifacts, such as proposals, strategy documents,
course materials etc, may be captured in the context
of the activities that require or produce them.
Concepts, such as monitoring and evaluation and
response to events (see Figure 5) are often best
represented in Class Diagrams where the elements
concerned and their interactions can be depicted.
The business rules such as those for determining the
appropriate response to events or for decisions in
workflows are captured as constraints.
4 CHECKLISTS
Checklists are the result of a non-formal synthesis of
knowledge of the domain. Tables 1 and 2 based on
Hirumi (2003), illustrate the knowledge areas some
of the widely used checklists represent and show the
variety in scope and nature of the checklist areas.
These lists were developed by a variety of processes,
few of which were fully documented but include
surveys of practice, expert submissions, team
brainstorming and formalizations of working
practices. Using these major sources, a composite
list was developed for some aspects of the e-learning
development process. A sample of the composite is
presented in Table 2. The style of checks varies
significantly. Some are checks that represent points
of principle, some are on approach, some on
activities undertaken and some are instructions about
what to do. The sample in Table 2, and the type used
in this study are of the style that relate to activities
undertaken and objectives achieved. In the
composite checklist an attempt has been made to
keep consistent checks that relate to activities and
objectives.
The process of consolidating the various
checklists consists of an iterative amalgamation and
breakdown of the various activities represented by
the checks. By iteratively cross-checking checks it is
possible gradually to extend the scope of the
subjects checked and to avoid repetition. By
CROSS-DOMAIN MAPPING: QUALITY ASSURANCE AND E-LEARNING PROVISION
201
iteratively considering groups of checks it is
possible, on first principles, to assess the
completeness of the scope and the continuity of the
processes.
Table 2: Sample Section from Checklist for E-Learning
Development, University of Manchester
QA Checklist for Project Management
Pre-Planning
Has a structured approach been adopted?
Have roles and responsibilities been defined?
Has a communication protocol been agreed?
Has documentation been agreed?
Project Control
Does the project have an external assessor?
Has an evaluation, monitoring and feedback
system been set up?
Do you have a system for Change
Management?
Project Exit
Have the deliverables been accepted?
Have you decided how to measure whether the
deliverables
have been achieved?
Are there any remaining to be achieved at a
later stage?
How will you assess what lessons have been
learnt?
How will the final costs be calculated?
How will you assess if the benefits have been
achieved?
Also by iterative composition, and based on cues in
the original checklists, it is possible to develop a
structure to the checking process that represents
stages or components of a viable e-learning
development process. For each of the checks and
stages it is possible from some of the checks and
from first principles to associate actions and artifacts
elsewhere in the enterprise.
5 CROSS-DOMAIN MAPPING
There exists a two-way interaction between the QA
domain and the e-learning provision domain that
may also be captured by the evolving model.
Processes in these two domains interact with each
other and influence each other. QA may be viewed
as an “Aspect” of the e-learning domain that may be
modeled as a system running alongside and
impacting on the e-learning provision system.
In turn, as the domain model for e-learning
provision evolves, new checks will be discovered
that may be added to the checks repository. Gaps not
covered by checks may be identified in the QA
process and redundancies may be highlighted. Other
factors such as which checks are critical and whether
there is any bias in the checks may also be
illuminated by the act of modeling the e-learning
provision domain and capturing practices.
Checklists tell us which things in each stage of a
business process, activities and their artifacts should
be monitored, and what the outcomes of the checks
should be in terms of what actions should be taken
and what changes made if the results do not meet
quality criteria. This information allows us to build a
model of the business process itself.
A small worked example of this cross-domain
mapping process is given here using checks
available from an internal source (Petch, 2003) and a
few external sources (Frances and Bonora, 2004,
Kelly, 2004, QAA 1999). Figure 2 shows the top
level activity diagram for one section of the e-
learning provision model process.
Figure 2: Activity Diagram for Preparing a New Course
Proposal for Review.
This section covers the stages between a faculty
board approving a preliminary proposal for a new
course and requesting a detailed “New Course
Proposal” in order to execute a “New Course
Review” and the New Course Proposal being ready
for that review. The group (role) responsible for
carrying out these activities is referred to as the
“Course Team”. A checklist appropriate for this
stage in e-learning provisioning provided the
knowledge about the existence of the role of an
approving body. In many institutions this would be a
faculty board but in others it may not. In the latter
case the checklist may be indicating what roles
ICEIS 2005 - SPECIAL SESSION ON EFFICACY OF E-LEARNING SYSTEMS
202
another body may have to take on in order to carry
out the approval process. The checklists have also
guided the sequencing of the steps and in some cases
contain the prerequisites for activities.
The checks from multiple lists are managed in a
repository where they are given a logical
organisation based on the 15 identified practice areas
within the e-learning lifecycle (Dexter and Petch,
2003).
Figure 3: The Practice Areas in the e-Learning Lifecycle
Activities from these practices are executed at
various times during the lifecycle of a product such
as an “e-learning course”. For management
purposes, the whole lifecycle is divided into phases
and each phase is divided into a number of iterations
depending on the complexity of the product being
developed. In each iteration there are a number of
activities from the practices and the iteration
produces a set of deliverables.
The checklist items for the activity Market
Analysis, from the Activity Diagram in Figure 2, are
found in the “Business Analysis and Planning
practice (Figure 3). Checks were modeled as Classes
and Figure 4 shows the internal structure of a check
(attributes and operations) and its relationship to the
e-learning lifecycle.
There are two ways to build on the e-learning
domain model from the checklists:
1. Adding a hierarchy of activities that matches
the checklist items by using subactivity states,
drilling down from the top level activity
diagram and adding object flow states to link
artifacts (documents, software applications, e-
learning materials, technology) to the
activities. These are artifacts required or
produced by the activities.
2. Creating a Use Case for the activity. Each Use
Case may then be expanded to describe the
workflow and outputs in detail. The Use Case
will also specify its preconditions, i.e. the
activities that have to have been completed
prior to its execution. Each Use Case may then
be expanded to describe the workflow and the
outputs in detail.
Figure 4: Structure of Checklist Item
The following table (Table 3) shows the
activities discovered in checklists relating to
“Market Analysis” that would be relevant to the
stage in the process shown above, “Preparing New
Course Proposal”.
Table 3: Activities Identified for Market Analysis
Market Analysis Activity
Determine brand identity
Identify markets and the elearning segments
Determine the positioning of the course
Calculate the size of potential markets
Assess trends in potential markets
Assess ease of access to potential markets
Assess the nature of the competition
Determine the market share of other producers.
Create strategy for acquiring and analyzing market
information
Set up system for monitoring and evaluating needs of
students and alumni
Determine the long-term potential of the course
Identify the sales channels for the course.
Determine whether price is a determining factor
Discover which courses have done well recently and
why (also poorly)
Review possible changes in government policy that
may affect demand
Discover the key success factors in this market
When the course team reaches the stage in the
preparation of the New Course Proposal of “Market
Analysis” it will be able to see the expanded set of
activities recommended. The team should execute
the activities and then use a checklist from the
CROSS-DOMAIN MAPPING: QUALITY ASSURANCE AND E-LEARNING PROVISION
203
Figure 5: Events and Rules Governing Processes
repository to ensure that it has covered all the areas.
The relevant checks are found in the checklist
repository by the activity “Market Analysis” itself,
by means of a subscription mechanism (see section
6). Each of the activities inside “Market Analysis”
may also have subscribed to checks and these can be
made available to the course team as they execute
the activity Operationalization.
The model shown above (Figure 5) of the event
response governing process is based on a simplified
version of the event-driven process and data-event-
driven process models provided in the EDOC UML
profile (OMG, 2004).
This model decouples the events generated by
an activity or artifact in the business process from
the set of responses to that event by using a publish-
and-subscribe mechanism. In this way any activity
or artifact in the system can subscribe to a set of
checks and respond to them appropriately. Any
event in the system, generated by an activity or an
artifact can publish, in an event notice, the need for a
set of checks and these will be picked up by those
processes that have subscribed to the event. Their
response to the checks is contained in the “Response
Rule”. This response could be in the form of a new
set of activities in a process and/or the repetition of
activities that have already been executed.
This mechanism allows a reservoir of checklist
items to serve multiple processes, with checklist
items being used in different places in ways
determined by the Response Rule which will be
appropriate for the context.
6 E-LEARNING SERVICES
PROVISION
e-Learning service provision can be driven by
executable business process models by adopting
mechanisms based on the Business Process
Execution Language for Web Services (BPEL4WS)
(Kath, Blazarenas, Born, Eckert, Funabashi and
Hirai, 2004). Such mechanisms will collect the
services and components in the environment, both
technology-based and people-based, and
choreograph them into a service aligned with the
defined task.
In order to get closer to the quality of model
required for such a venture we need to be able to
acquire extensive, in-depth knowledge about the
processes. The models must also be provided to
institutions in a way that they may be customized for
the organization. One means to improve the depth of
knowledge in the domain model is shown to be by
interacting with the QA domain and to learn from
QA checklists.
7 CONCLUSION
We have argued that cross-domain mapping can
form part of an enterprise modeling strategy for e-
learning provision. We have demonstrated a proof of
concept for the process of cross-domain mapping
and have provided a model of a mechanism to
operationalize the use of checklists for governing the
provisioning process in e-learning.
ICEIS 2005 - SPECIAL SESSION ON EFFICACY OF E-LEARNING SYSTEMS
204
The next steps in the work are a fuller
development from the proof of concept to a rich e-
learning model based on the combined checklist set.
At the same time, the checklists should be refined as
we are able to take a whole system view. The
checklist operationalization mechanism should be
implemented and tested over a range of situations
where checks give rise to modified activities in
multiple parts of the e-learning provision processes.
REFERENCES
Dexter, H., Petch, J., 2003. A Roadmap for e-Learning
Service Provisioning, e-Learning Research Centre,
research paper,
http://www.elrc.ac.uk/publications/163/
Frances, V.L., Bonora, A.G., 2002. Methodology for the
Analysis of Quality of Open and Distance Learning
Delivered via the Internet (MECA-ODL)
www.adeit.uv.es/mecaodl
Franklin, T, Petch, J., Armstrong, J., Oliver M., 2004. An
Effective Framework for the Evaluation of E-
Learning, eLearning Research Centre, research paper
http://www.elrc.ac.uk
/publications/153/
Fuentes, L., Vallecillo, A., 2004. An Introduction to UML
Profiles, UPGRADE , European Journal for the
Informatics Professional 5,2,
http://www.upgrade-
cepis.org/issues/2004/2/
upgrade-vol-V-2.html
Hirumi, A., 2003. In Search for Quality: An Analysis of e-
Learning Guidelines and Specifications, submitted to
the Journal of e-Learning Administration
International DOI (Digital Object Identifier) Foundation,
2005. Glossary,
http://www.doi.org/
handbook_2000/glossary.html
Kath, O., Blazarenas, A., Born, M., Eckert, K.-P.,
Funabashi, M., Hirai, C., 2004. Towards Executable
Models: Transforming EDOC Behavior Models to
CORBA and BPEL, Proceedings of the 8th IEEE Intl.
Enterprise Distributed Object Computing Conference
Kelly, B., 2004. QA Focus Handbook, JISC
QA Focus Project
www.ukoln.ac.uk/qa-focus/
documents/handbook
Object management Group (OMG), 2004. UML Profile
for Enterprise Distributed Object Computing
Specification,
http://www.omg.org/
technology/documents/
modeling_spec_catalog.htm
Petch, J., 2003. Quality Assurance for New Courses, e-
Learning Research Centre, working paper,
http://www.elrc.ac.uk/publications/162/
Petch, J., 2004. QA Checklists, e-Learning Research
Centre, briefing paper, http://www.elrc.ac.uk
/publications/162/
Quality Assurance Agency for Higher Education (QAA),
1999. QA Distance Learning Guidelines
www.qaa.ac.uk/public/dlg/contents.htm
Scienter-MENON Network (ed), 2004. Sustainable
Environment for the Evaluation of Quality in E-
Learning (SEEQUEL)
http://www.education-
observatories.net/seequel/SEEQUELTQM
_Guide_for_informal_learning.pdf
Western Cooperative for Educational Telecommunications
(WCET), 2000. Guidelines for the Evaluation of
Electronically Offered Degree and Certificate
Programs
http://www.wcet.info/resources/accreditation/
guidelines.pdf
Wilcox, P., Petch, J., Dexter, H, 2004. A Foundation for
Modelling e-Learning Processes, e-Learning Research
Centre, working paper,
http://www.elrc.ac.uk/publications/148/
CROSS-DOMAIN MAPPING: QUALITY ASSURANCE AND E-LEARNING PROVISION
205