A FRAMEWORK FOR DEVELOPMENT AND MANAGEMENT
OF E-LESSONS IN E-LEARNING
Azita A. Bahrami
Department of Information Technology
Armstrong Atlantic State University
11935 Abercorn Street
Savannah, GA 31419
Keywords: E-lesson’s model, E-lesson Cube, E-lesson’s View, E-learning, and Ontology.
Abstract The use and or re-use of the existing e-lessons for the creation of new ones make the e-learning both time
and cost effective. To accomplish this, however, requires the removal of some obstacles first. This paper
presents a framework for that purpose. The progression of the concepts leading to the framework includes
the introduction of a multi-dimensional e-lesson model that leads to the construction of an e-lesson cube.
This cube is the backbone of an e-lesson warehouse, which in turn is the main component of the proposed
framework.
1 INTRODUCTION
In general, any “lesson”, λ includes a subject, S
(topic), designed for a specific audience, A, in a
specific field, F, of study utilizing a delivery system,
D, for conveyance of the lesson. This can be
expressed as λ = (S, A, F, D) with the subject’s
value of ω = {o
1
, o
2
, . . ., o
p
}, where os are a set of
related and organized objects (sub-topics). This
makes a lesson a four dimensional object which
carries the value of ω.
An e-lesson, however, is only three
dimensional because there is only one and the same
delivery system for all e-lessons, namely, the
Internet. As a result, an e-lesson can be viewed as a
3-D object, λ = (S, A, F), with a value composed of
a set of digitized objects. For example, an e-lesson
teaching the Subject “DNA” to the audience of
university freshman students majoring in the field of
biology and using teaching objects of enzymes,
chromosomes, and genes would be noted in the
following general form,
λ
1
=(“DNA”, “University Freshman”, “Biology”)
and
ω
1
= {Enzymes, Chromosomes, Gene}.
Each digital object in ω is an integral part of the e-
lesson and may be composed of digitized text, data,
image, sound, video, etc. or some combinations of
them.
In the e-learning arena, DNA, for example, can
be taught to the following groups of students: (1)
university freshman in biology, (2) Ph.D. students in
biology, (3) university seniors in computer science,
and (4) university seniors in forensic medicine
(Table 1.) Naturally, there would be four different
sets of values for the e-lessons suitable for these
groups encompassing different breadth, depth, and
emphasis. In explaining the point, suffice it to say
that a university freshman in the field of biology
may need the general and basic ideas about DNA
whereas a Ph.D. student of the same field may need
an in-depth study of DNA for the purpose of
acquiring deep knowledge of DNA’s functions. A
university senior in the field of computer science
studying DNA, however, may need to learn not the
DNA’s functions but the replication of DNA strands
and the “patterns” that they carry. And lastly, a
university senior student in the field of forensic
medicine may study DNA from the view point of
criminal investigations.
504
A. Bahrami A. (2005).
A FRAMEWORK FOR DEVELOPMENT AND MANAGEMENT OF E-LESSONS IN E-LEARNING.
In Proceedings of the First International Conference on Web Information Systems and Technologies, pages 504-509
DOI: 10.5220/0001226705040509
Copyright
c
 SciTePress

Table 1: A set of e-lessons.
λ
1
= (“DNA”, “Univ. Freshman”, “Biology”)
λ
2
= (“DNA”, “Univ. PhD students”, “Biology”)
λ
3
= (“DNA”, “Univ. Seniors”, “Computer
Science”)
λ
4
= (“DNA”, “Univ. Seniors”, “Forensic
Medicine”)
Let us assume that the four versions of the e-
lesson on DNA in Table 1 are already developed and
in use. Now the development of a new e-lesson on
the same subject for the freshman students of
medicine is being attempted. Can some parts of the
values
from the existing four e-lessons be used for
this fifth group of students? To answer this
question, it would be beneficial to first lay out the
ground work for an e-lesson development.
• Time and Cost. The most time consuming
and costly part of an e-lesson development is
the creation of digital content for an e-lesson
because an e-lesson requires the valuable
time of a field/domain expert in both the
initial design-and-development and the
subsequent modifications and or re-writing of
the lesson for different audience in the same
or different fields. Therefore, using the
existing e-lessons whose glitches, for the
most part, are already removed can shorten
and lessen the amount of time and cost
necessary for development of a new e-lesson.
Though the original e-lesson
developer can be consulted, if necessary, for
any modification, the required time and cost
would still be less than developing an e-
lesson from ground zero.
Furthermore, providing links to the
existing e-lessons inside the new e-lesson can
substantially shorten the e-lesson itself or,
rather, expand the e-lesson beyond the
boundaries of the intended objectives of the
present lesson.
• Success Rate. Using the existing e-lessons
or parts of them whose content validities
have inevitably been already tested would
improve the chances of success for the new
e-lesson, and thus the e-learning experience.
To sum up the points made, it may be stated that an
e-lesson can be constructed time-and-cost effectively
through using, re-using, and referencing the existing
lessons. To actualize such idea, however, requires
the removal of some obstacles first. The obstacles
are: (a) how to find the e-lessons of interest and (b)
how to utilize the found e-lessons.
An architecture has been presented by Siqueira
et al (Siqueira et al, 2002) that resolves the first
obstacle. This paper, however, presents a
framework for the removal of the both obstacles
making the use, re-use, and reference to other
existing lessons possible. The foundation of the
framework is investigated in section 2. The
framework itself and the relevant discussion are
covered in section 3. The conclusion and future
research are included in section 4.
2 THE FOUNDATION OF THE
FRAMEWORK
An e-lesson has the following properties:
(a) Dimensions. An e-lesson is a 3D object,
λ = (S, A, F).
(b) Value. An e-lesson has a value,
ω = {o
1
, o
2
, . . ., o
p
}.
(c) Owner. The owner of an e-lesson has the right
to his/her intellectual property and is the
authority for granting permission and profiting
from the lease of his/her property, if so he/she
chooses.
(d) Host. A host is the computer in which the e-
lesson resides.
(e) View. The information, metadata, regarding an
e-lesson, which is within the “content” section
of a meta tag inside the HTML source, is
considered to be the view of the e-lesson. The
metadata contains the actual descriptive-words
about the value of the e-lesson including
keywords, terms, phrases, etc. The structure of
metadata for e-lessons is aimed to become
standardized (Anderson et al, 1999, Hermans et
al, 1999, and Hodgines et al 2002) and the EU
commission initiative on e-learning and IEEE
are the two major forces behind this effort. In
order for e-lesson developers to create a new e-
lesson, they need to query the views, manipulate
the views, and map the views, all of which are
addressed below.
2.1 Querying the Views
An e-lesson designer needs to query metadata of
existing e-lessons of interest related to the e-lessons
A FRAMEWORK FOR DEVELOPMENT AND MANAGEMENT OF E-LESSONS IN E-LEARNING
505

he/she intends to develop. As the result of not yet
having a set of standard in place for metadata, some
relevant information fail to return. An example of
such case is when the designer queries about “ER
Model” and may not have any return because the
existing e-lessons may have presented their
descriptive-words for the same concept as “E-R
Model”, “Entity Relationship Model”, “Database
Top-down Model” or other variations.
Quite clearly, this is a problem needing to be
solved. The solution to this major problem is to
create ontology. The ontology would represent the
metadata items of e-lessons and their relationships
with one another in a thorough and formal fashion.
In other words, ontology is more than a multi-
faceted taxonomy because it includes the
descriptive-words of the metadata, their complex
relationships, and the rules about how to specify
descriptive-words and relationships (Neches et al,
1991). One of the relationships that is established
among the metadata items from the view point of
ontology is the “concept hierarchy”. For example,
an e-lesson designer looks for a descriptive-word,
such as “Database Design” and receives no return
because none of the existing e-lessons contain these
descriptive-words in its meta tag. Yet, the
descriptive-words such as “Universal Relation
Model” and “ER Model” that are two well known
approaches in “Database Design” could be found.
The hierarchical relationship among the three
descriptive-words “Universal Relation Model”, “ER
Model” and “Database Design” is identified and
kept in the ontology that includes these three
descriptive-words. Naturally, due to the large
number of possible values for each dimension of the
e-lessons and their combinations, there is a need for
a library of ontologies along with a library interface.
2.2 Manipulating the Views
To develop a new e-lesson, the developer needs to
group and re-group the existing e-lessons along the
three dimensions of S, A, and F in various
combinations. This can be accomplished through
“e-lesson cube” along with the needed set of
operations.
2.2.1 E-lesson Cube and Needed Operations
A set of existing e-lessons and their values are
given, Table 2. The assumption is that the set of
possible values for dimension audience is: university
freshman students (UF), university sophomore
students (USP), university junior students (UJ),
university senior students (US), university master
students (MS), and university Ph.D. students (PhD).
Also, the set of values for the field dimension is:
business administration (BA), which includes both
computer information systems (CIS) and
management (MG), computational sciences (COS),
which includes information technology (IT) and
computer science (CS), and electrical engineering
(EE), which includes computer engineering (CE),
electronics (EL), and power (EP).
Table 2: A set of existing e-lessons and their values.
λ
1
= (“ER Model”, “MS”, “CIS”)
ω
1
= {o
1
1
, o
1
2
, . . ., o
1
p1
}
λ
2
= (“ER Model”, “US”, “IT”)
ω
2
= {o
2
1
, o
2
2
, . . ., o
2
p2
}
λ
3
= (“ER Model”, “US”, “CS”)
ω
3
= {o
3
1
, o
3
2
, . . ., o
3
p3
}
λ
4
= (“ER Model”, “US”, “CS”)
ω
4
= {o
4
1
, o
4
2
, . . ., o
4
p4
}
λ
5
= (“Router”, “UJ”, “CS”)
ω
5
= {o
5
1
, o
5
2
, . . ., o
5
p5
}
λ
6
= (“Router”, “USP”, “CE”)
ω
6
= {o
6
1
, o
6
2
, . . ., o
6
p6
}
λ
7
= (“Channels”, “MS”, “IT”)
ω
7
= {o
7
1
, o
7
2
, . . ., o
7
p7
}
λ
8
= (“Universal Relation Model”, “UJ”, CS)
ω
8
= {o
8
1
, o
8
2
, . . ., o
8
p8
}
λ
9
= (“Universal Relation Model”, “US”, “MG”)
ω
9
= {o
9
1
, o
9
2
, . . ., o
9
p9
}
Let us only consider those lessons for which
the subject is “ER Model” (i.e. λ
1
, λ
2
, λ
3
, and λ
4
.)
Since these e-lessons have the same subject, they
may be
WEBIST 2005 - E-LEARNING
506

Figure 1: E-lesson Cube Formation: (a) 2-D model for those lessons of Table 2 with subject “ER Model”, (b) 2-D model for
those lessons of Table 2 with the subject “Router”, and (c) the e-lesson cub for the lessons of Table 2.
shown in a two dimensional space, a table, Figure
1.a. The e-lessons for the subject “Router” (i.e. λ
5
and λ
6
) are also shown in the table of Figure 1.b.
Thus, for each subject a table can be produced.
These tables collectively make an e-lesson cube,
Figure 1.c. As a result, an e-lesson cube has three
dimensions and a collection of values.
The question is through what specific
operations the grouping and regrouping of the
existing e-lessons are accomplished? The needed
operations are as follow.
Roll-up operation. The values for any of the
dimensions (S, A, F) can be collapsed to make a
new value within the “concept hierarchy”. For
example, the values on the subject’s dimension of
the e-lesson cube of Figure 1.c can be collapsed
into higher level subjects of “Database Design”
and “Network”. As a result, the tables for the
subjects of “Universal Relation Model” and “E-R
Model” are integrated to make a new table for the
“Database Design” subject. The tables for the
subjects of “Router” and “Channel” are also
integrated to make the table for the “Network”
subject.
Roll-down operation. The opposite of roll-up
operation can be achieved by expanding one or
more dimensions. For example, field’s dimension
may expand from (COS, BA, and EE) to (CS, IT,
CIS, MG, CE, EL, and EP).
Slice operation. The e-lesson cube can be cut along
only one of the three dimensions for a set of values
on that dimension.
Dice operation. The e-lesson cube can be cut along
more than one dimension for a set of values (on
those dimensions). Dicing the e-lesson cube of
Figure 1.c for Subject = (ER Model | Universal
Relation Model) and Audience = (US | MS) creates
a sub-cube for which S =( ER Model, Universal
Relation Model), A = (US, MS), and F = (all the
fields)).
Pivot operation. Any e-lesson cub, or results
produced by any of the above operations can be
rotated so that each dimension may become the
focal point of observation.
The above operations can be delivered
through a technology called On-line Analytical
ω
1
UF USP UJ US MS PhD
ω
2
,
ω
3
,
ω
4
A u d i e n c e
COS
BA
EE
F
i
e
l
d
(a) (b)
COS
BA
EE
ω
5
ω
6
UF USP UJ US MS PhD
F
i
e
l
d
A u d i e n c e
S u b j e c t
(C)
A u d i e n c e
Router
Channel
Univ. Rel. Model
ER Model
COS
BA
EE
UF USP UJ US MS PhD
ω
1
ω
5
F
i
e
l
d
ω
2
,
ω
3
,
ω
4
A FRAMEWORK FOR DEVELOPMENT AND MANAGEMENT OF E-LESSONS IN E-LEARNING
507

Processing (OLAP) (Han et al, 2001, and
Chaudhuri, 1997)
. For the ease of manipulation,
the exiting e-lessons available to a developer may
be stored in a data warehouse. Data warehouses
are specifically designed and developed for multi-
dimensional data and are used by OLAP
technology. Therefore, a data warehouse is quite
suitable for creating, storing, and manipulating the
e-lesson cube model proposed in this paper.
2.3 Mapping the Views
The outcome of applying a group of operations on
a warehoused e-lessons is a very large set of
values for multiple e-lessons. The question is that
how such a large volume of data may be used by a
designer. To answer this question, it can be
assumed that the set of lessons’ values Ω = {ω
1
,
ω
2
. . ., ω
n
} belonging to the set of e-lessons L =
{λ
1
, λ
2
, . . . , λ
n
} is delivered by OLAP. Let the
union of the digital objects of the e-lessons’ values
in Ω be O = {o
a
, . . ., o
q
} and the union of the
views (descriptive-words) for the lessons in L be
W = {w
e
, . . ., w
m
}. There is a many-to-many
relationship between the elements in O and W,
which can be easily broken into two 1:n
relationships and the results be stored in a
relational database for the actual use by the
designer.
3 THE FRAMEWORK
The proposed framework has six components
shown in Figure 2. These components are, “E-
lesson Warehouse”, “OLAP Technology”,
“Library of Ontologies”, “Library Interface”
“Mapper”, and “Controller”. Each component is
briefly described below.
E-lesson Warehouse. This is a warehouse of e-
lessons collected by the designer composed of a set
of smaller warehouses (data marts), each
containing e-lessons of one subject only. The e-
lessons of the warehouse are integrated from many
heterogeneous
data repositories that are either structured
(databases) or unstructured (flat files) and are
distributed.
OLAP Technology. This technology is capable of
handling the operations “roll-up”, “roll-down”,
“slice”, “dice”, and “pivot”. For details of OLAP
technology consult (Han et al, 2001, and
Chaudhuri, 1997)
Library of Ontologies. This is a collection of
ontologies used for different subjects in the subject
domain.
Library Interface. This component has a dual
function: (a) It helps the Controller by providing
the descriptive-words related to the ones submitted
by the designer through an OLAP command and
(b) it facilitates the library’s maintenance.
Mapper. This component provides a detailed
mapping of the views and values on each other,
using the sub-component Indexer, and stores the
findings in a relational database. The creation,
manipulation and maintenance of this database are
also the function of the Mapper.
Controller. This is the component that (a)
receives a query from the user, (b) examines the
descriptive-words in the query for possible
replacement using the “library of ontologies”, (c)
re-writing the user query, if it is necessary, and (d)
Obtaining the answer to the query and returning it
to the user. The Controller can also provide the
user with any mapped views and values.
It was previously mentioned that to make the
e-learning more time and cost effective, it is highly
desirable to use, re-use, and reference existing e-
lessons. It was also indicated that such an attempt
requires the removal of two obstacles: (a) how to
find the e-lessons of interest and (b) how to utilize
the found e-lessons. The proposed framework of
Figure 2, addresses both obstacles with solutions.
To justify the solutions, let us discuss how the
proposed framework provides for each obstacle.
How to find the e-lessons of interest. A designer
wants to find all the e-lessons that teach, for
example, the “Database Design” subject to a
freshman audience. As soon as the “Controller”
receives this request from the user, it consults the
“library of ontologies” through the “library
interface”. The library determines the fact that the
WEBIST 2005 - E-LEARNING
508

“Database Design” subject is located in a. higher
Figure 2: The Framework
level of “concept hierarchy” and is composed of a
set of lower level concepts {ER model, Universal
Relation model}. This set of concepts is
communicated back to the Controller by the
“library interface”. The Controller modifies the
user’s query to include all three descriptive words
“Database Design”, ER model, and Universal
Relation model. The new query is implemented by
the OLAP that finds all the e-lessons of interest
and returns them to the user through Controller.
How to utilize the found e-lessons. The mapping
of the views onto the e-lessons’ values, by the
Mapper component of the proposed framework,
makes it possible for the designer to use the e-
lessons’ values in a constructive way. The creation
and maintenance of a database for the mapped
values enable the designer to retrieve any
relationship between the views and the values on-
line. In fact, the Mapper provides a powerful filter
for isolating the needed materials among high
volumes of the e-lessons’ values delivered by an
OLAP operation.
4 CONCLUSION AND FUTURE
RESEARCH
The proposed framework makes it possible for a
designer to use, re-use and or reference the existing
e-lessons. The great benefit of using the
framework is that the time and cost of development
a new e-lesson lowers, making the entire e-learning
experience time and cost effective. The
implementation of the proposed framework and the
investigation of its behavior are the focus of the
future research.
REFERENCES
Siqueira S. W. M., Braz M. H. L. B., and Melo R. N.,
“E-Learning Content Warehouse Architecture”,
Proceedings of the International Conference of
WWW/Internet, pp. 739-742, Nov. 2002, Lisbon,
Portugal.
Anderson, T, McArthur D., Griffin S., and Wason T.,
“IMS Learner Resource Meta-data Best Practices
and Implementation Guide v.1”, Revision: August
1999,
http://www.imsglobal.org/metadata/mdbest01.html.
Hermans, H. J. H., Koper E. J. R., Loeffen A.,
Manderveld J. M., and Rusman E.M., “Reference
Manual for Edubox-EML/XML binding 1.0/1.0
(Beta Version)”,
http://dspace.learningnetworks.org/retrieve/122/EML1.0
-ReferenceManual.pdf
Hodgins W., (Chair), Duval E. (Technical Editor),
“Draft Standard for Learning Object Metadata IEEE
P1484.12.1/D6.4”, July 2002,
http://ltsc.ieee.org/wg12/files/LOM_1484_12_1_v1_
Final_Draft.pdf
Neches R., Fikes R., Finn T., Gruber T., Patil R., Senator
T., and Swartout W.R., “Enabling technology for
knowledge sharing”, AI Magazine, Vol. 12 , Issue 3,
pp. 36-56, 1991.
Han J., Kamber M., “Data Mining: Concepts and
Techniques”, Morgan Kaufmann Publishers, 2001.
Chaudhuri S. and Dayal U., “An overview of Data
Warehousing and OLAP Technology”, ACM
SIGMOD Records, No. 26, pp. 65-74, 1997.
OLAP Technology
U
S
E
R
(Designer)
Library
Interface
Library of
Ontologies
. . .
E-Lessons Warehouse
Controller
Mapper
Database Indexer
Data-Mart 1 Data-Mart n
A FRAMEWORK FOR DEVELOPMENT AND MANAGEMENT OF E-LESSONS IN E-LEARNING
509