Context Metadata for e-Learning Environments (Richter 2007)
1. KGIT
Korean German Institute of Technology
e-Learning Research Center
Universität Duisburg-Essen
Fachbereich Wirtschaftswissenschaften
Project Research Report for the Project
“Context Metadata for e-Learning Environments”
Research period: 2007 February - 2007 August
Author: Thomas Richter, Senior Researcher
Purpose: This report shows the recent state of research of the finalized project-step. It
shall finally lead to a standardization process, which Jan will try to start at next DIN –
meeting in October.
Recipients: Dr. Han Tae In
Dr. Jan M. Pawlowski
2. Abstract:
Title: A metadata specification for the context of e-Learning environments
Author/Date: Thomas Richter, 2007 July
Task (Description):
The aim of this document is an approach to systematically specify potential candidates for
context-metadata in e-learning environments.
Metadata in general within the surrounding of e-learning environments are a necessity, at
least to realize i.e. a successful search for a course or course-modules which have to fit
specific needs (i.e. subject, language, etc.). They also can provide information on technical
environments needed to get use of the course, on knowledge demands or recommendations
to understand it, on used didactical and instructional approaches within the courses, and
others. The context of e-Learning, which plays a significant role when it comes to adapta-tion
processes within the internationalization of contents is not part of the already stand-ardized
metadata specifications (Learning Object Metadata [IEEE02], DIN-PAS 1032-2
[DIN04b], a.o.), which primarily focus on procedures and attributes of courses. Defining
the context of e-Learning with metadata might not be possible to be fulfilled in complete-ness,
because it is a growing process, but a basic system can be defined to prepare a better
compatibility between course-modules and to avoid greater problems (i.e. synonyms,
homonyms) and as a discussion basis for standardization efforts in later times.
A view on existing specifications particularly on their interfaces to bind in those defined
metadata is obligatory. The necessity of adaptations will have to be discussed as well as
first approaches to describe such procedures will be defined. A critical view concerning the
subject and recommendations for further research will round off the document.
3. Content
1 Introduction ............................................................................................................................................... 1
2 Fundamentals............................................................................................................................................. 2
2.1 e-Learning............................................................................................................................................2
2.1.1 e-Learning Environment (Definition)........................................................................................3
2.1.2 e-Learning Context (Definition)................................................................................................3
2.1.3 e-Learning Context-Metadata (Definition)................................................................................3
2.1.4 Adaptation of e-Learning (Definition) ......................................................................................4
2.2 Standards - a Brief Overview ..............................................................................................................4
2.2.1 Learning Design (LD) ...............................................................................................................5
2.2.2 Learning Objects and Learning Objects Metadata (LOM)........................................................6
2.2.3 Metadata for Learning Resources (MLR) .................................................................................8
2.2.4 DINI / ELAN: Metadaten für elektronische Lehr- und Lernmaterialien ..................................9
2.2.5 Dublin Core Metadata Schema..................................................................................................9
2.2.6 DIN-PAS 1032-1 .....................................................................................................................10
2.2.7 DIN-PAS 1032-2 .....................................................................................................................10
2.2.8 AICC – Aviation Industry Metadata Specification (AIMS)....................................................10
2.2.9 Sharable Content Object Reference Model (SCORM) ...........................................................11
2.2.10 Intermediary Conclusions........................................................................................................11
2.3 The Context of e-Learning (Brief Description) .................................................................................12
3 Methods to Determine Metadata ............................................................................................................. 14
4 The Context of e-Learning ...................................................................................................................... 16
4.1 Influence Factors on e-Learning........................................................................................................17
4.2 Contextual Changes and Dynamics ...................................................................................................19
4.2.1 Using the “Influences Time Delay Model” (ITDM): Concrete Example ...............................20
4.2.2 Evaluation of the ITDM for Further Use.................................................................................24
4.2.3 Context-metadata and the Contextual Environment................................................................25
5 Culture ..................................................................................................................................................... 26
5.1 Culture Description and Comparison Models ...................................................................................28
5.1.1 HOFSTEDE’S “Onion-Model” related to the e-Learning Scenario ...........................................28
5.1.2 HOFSTEDE’S “Dimensions of Culture” related to the e-Learning Scenario.............................30
5.1.2.1 Evaluation of the described Culture Classification Model ............................................ 41
5.1.3 The Model of TROMPENAAR’S and HAMPDEN-TURNER..........................................................45
5.1.4 The 3-Cultural-Dimensions Model of E. T. HALL ..................................................................47
5.1.5 The 14-Dimensions Model of HENDERSON.............................................................................48
5.2 Conclusions and Metadata Definition ...............................................................................................51
6 Determination of Context-Metadata for e-Learning Courses and Environments ................................... 55
6.1 Culture and the e-Learning Context ..................................................................................................56
I
4. 6.1.1 Language .................................................................................................................................56
6.1.2 Cultural Differences in Communication .................................................................................57
6.1.3 Humor......................................................................................................................................58
6.1.4 Cultural Specific Use of Media and Acceptance Level...........................................................58
6.1.5 Cultural Related Gender Differences ......................................................................................59
6.1.6 Social Capital...........................................................................................................................60
6.1.7 Habits and Preferences ............................................................................................................61
6.1.8 Acceptance of Technology and Knowledge............................................................................62
6.1.9 Society’s General Opinion ......................................................................................................62
6.1.10 Taste 62
6.1.11 Indigenous Cultures.................................................................................................................63
6.1.12 Intercultural Hints for Technology Transfer ...........................................................................63
6.1.13 Pedagogical Approach.............................................................................................................65
6.1.14 Culture related Context Metadata, Summation of Chapter 6.1 ...............................................65
6.2 Demographical Development ............................................................................................................67
6.2.1 Demographic Development and Demographic Trends ...........................................................67
6.3 Religion .............................................................................................................................................69
6.4 Technical Infrastructure.....................................................................................................................70
6.5 Rights.................................................................................................................................................73
6.5.1 Acceptance by Government and Accreditation.......................................................................73
6.5.2 Intellectual Property Rights.....................................................................................................74
6.5.3 Data Protection Rights.............................................................................................................74
6.5.4 Specific Copyright...................................................................................................................74
6.5.5 Laws Concerning the Usage of Internet ..................................................................................74
6.5.6 Business Related Laws ............................................................................................................74
6.5.7 Other Laws inflicting e-Learning ............................................................................................74
6.6 History ...............................................................................................................................................75
6.7 Politics ...............................................................................................................................................78
6.7.1 Model to describe politics out of the view of e-Learning .......................................................78
6.7.2 Independence of each view from the others ............................................................................79
6.7.3 Relevance of each view for e-Learning...................................................................................80
6.7.4 Conclusion...............................................................................................................................81
6.8 State of Development ........................................................................................................................81
6.9 Media Richness..................................................................................................................................83
6.10 Financial Aspects...............................................................................................................................83
6.11 Human Actors....................................................................................................................................86
6.12 Rules: Regional and Company-Individual Standards, Specific Agreements ....................................93
6.13 Companies .........................................................................................................................................94
6.14 Geography and Educational Infrastructure........................................................................................98
6.15 Learner Satisfaction – Known Demands ...........................................................................................99
6.16 Internet Security in e-Learning........................................................................................................100
II
5. 6.17 Relation to other Standards..............................................................................................................102
6.18 Conclusion .......................................................................................................................................102
7 Context-metadata Shapes and a Possible Data-Structure...................................................................... 105
8 Procedure Model to Determine Changing Needs Using the Context-metadata for e-Learning
Environments......................................................................................................................................... 111
8.1 Determining the Context-Metadata for e-Learning and Procedure Model......................................116
8.2 Automated Collecting of Context-metadata for e-Learning............................................................119
9 Critical View ......................................................................................................................................... 122
10 Conclusion............................................................................................................................................. 126
11 Future Perspectives................................................................................................................................ 127
12 Ongoing Research, Planned Projects..................................................................................................... 128
Bibliography ........................................................................................................................................................ i
Annex A.............................................................................................................................................................. a
TABLES
Table 1: Changes in the system as follows on context related impacts ............................................................23
Table 2: Exemplary values for Power Distance Index (PDI)............................................................................33
Table 3: Exemplary Values for Individualism Index (IDV) .............................................................................36
Table 4: Exemplary values for Masculinity Index (MAS) ...............................................................................38
Table 5: Exemplary values for Uncertainly Avoidance Index (UAI)...............................................................40
Table 6: Exemplary values for Long-Term Orientation Index (LTO)..............................................................41
Table 7: Direct comparison of PDI, IDV and UDI ...........................................................................................43
Table 8: Summary of found context-metadata related to Chapter 5 .................................................................54
Table 9: Culture related Context Metadata, Summation of Chapter 6.1...........................................................66
Table 10: Context Metadata Definition, Summation of Chapters 6.2, 6.3, 6.4 ................................................72
Table 11: Context Metadata Definition, Summation of Chapters 6.5, 6.6 .......................................................77
Table 12: Context Metadata Definition, Summation of Chapters 6.7, 6.8, 6.9 and 6.10..................................85
Table 13: Context Metadata Definition, Summation of Chapters 6.11, 6.12 and 6.13.....................................97
Table 14: Context Metadata Definition, Summation of Chapters 6.14, 6.15 and 6.16...................................101
Table 15: Dependencies of Context Blocks and Related Influence Factors ...................................................108
Table 16: Context Metadata, Final Summation ...................................................................................................i
ILLUSTRATIONS
Illustration 1: e-Learning Scenario 1 ................................................................................................................18
Illustration 2: e-Learning Scenario after environmental impacts, 5 years after reunion...................................21
Illustration 3: “Reunion of Germany”: Event-flow and time-shifted impacts within the system.....................23
III
6. Illustration 4: The „Onion“: Manifestations of cultures at different levels of depth [HoHo05].......................28
Illustration 5: The influence factors on e-Learning environments....................................................................55
Illustration 6: Relationship of other standards to this document ....................................................................102
Illustration 7: Contextual influence factors on e-Learning environments and dependencies .........................103
Illustration 8: Dependencies between Contextual Blocks...............................................................................109
Illustration 9: The adaptation process [RiPa07]..............................................................................................111
Illustration 10: Procedures to gather and compare data and to evaluate changing needs...............................114
Illustration 11: Procedure Model for the Data Gathering Process ..................................................................118
Illustration 12: XML Form of Fischer Weltalmanach Data Set (cutout)........................................................120
Illustration 13: List Form of Fischer Weltalmanach Data Set (cutout)...........................................................120
IV
7.
8. 1 Introduction
When e-Learning came up to become public, it primary was used to realize a personally
motivated further education in a concrete subject in the workplace or at home. The aim was
to fit new job-related demands or to serve private interests. Besides this basic idea, which
is one aspect of the concept of lifelong learning (adult education), basic school education
and academic education are targeted, too. Maybe in the future (my claim), e-Learning
could be a solution to compensate the negative demographic development in Central Eu-rope
(decline of birth-rate) as well as the related closures and accumulations of basic
schools. The worth of e-Learning for basic education particularly could be very high in ar-eas
with a low population density, i.e. parts of Africa. The reusability of content is a basic
idea related to e-Learning to avoid unnecessary doubled work and lower costs, so that the
accessibility for this kind of education can be raised even for economically poor countries.
That for modularization is a crucial key concept. Dividing full courses into reusable single
standing modules allows a better compatibility. A course in this meaning is a composition
of single standing modules. Each module i.e. represents a chapter of a classical book or
even smaller unities (like single illustrations). Different to the book’s chapters, the e-
Learning modules can be reused in various courses even in those with very different
subjects. Modularisation not only can avoid unnecessary rewritings of courses to same sub-jects
for different environments, but also prevents some risks. One of the risks is a possible
lack of comparability of different points of view, which is a basic demand of the construc-tivisticly
oriented didactic model. To make such a modularization useful, interfaces, pro-cedures
and data structures have to be standardized.
The context of e-Learning plays a significant role, when it comes to an adaptation of con-tents
into different societies (i.e. countries). Changes in various fields may be needed be-fore
a course can be implemented. Context-metadata may i.e. represent special character-istics
of societies and can be used to determine changing needs by comparison of the origi-nating
with the intended context. In this paper, metadata representing the context of e-
Learning have to be found, named, defined and ordered as preparation for a possible later
standardization. Further on, methods have to be found to gather those data and first ap-proaches
for procedures will be defined which show how to use the context-metadata
within the adaptation process.
As a result of the literature research it is to be found out if and which models to describe e-learning
context shall be used to support (ease) those procedures. Various context-metadata
are to be extracted out of the literature and determined through traceable considerations.
Next, a first logical order system has to be defined. After the context-metadata and an or-der
system have been determined and defined, first approaches of procedure models for the
adaptation process and the metadata determination process are to be presented. The com-patibility
of the context-metadata-system with the existing standards is to be checked and
maybe interfaces or possible adaptations have to be suggested. Methods to establish a
machine-aided collection of concrete context-metadata values are needed at least to excul-pate
authors from doing the definition work manually. It is to be proved how far an auto-matic
determination is possible and strategies are to be developed to collect the rest of the
data which cannot be collected automatically. Finally, a critical view on the matter of fact
itself and the way to solve the problems (open questions, hints) as well as remaining open
research needs have to be presented.
1
9. 2 Fundamentals
In this chapter, the basic information and the state of the art are documented: The subject
of this paper touches a lot of different fields of research. The research for this document
recommended taking literature of various fields into consideration, as Sociology, Psychol-ogy,
Information Technology, Economics, Anthroposophy, Pedagogies, Geography, Rights
and others. In a lot of cases, the literature itself gave the arguments to define topic-related
context-metadata. The literature in most cases has not directly been related to e-Learning,
so implications of the content led to the metadata. Demonstrating those argumentations
would force to cite those papers again, if already cited in the beginning of the paper or
otherwise by simply referencing former text in every single case, would lead to an unread-able
2
document.
An alternative approach and to avoid unnecessary doubling of information turning pages
around, special related literature is discussed to the beginning of and within each chapter
and only the common standards are discussed in Chapter 2.3. In the beginning of Chapter
5, the basics concerning the largest influencing factor (culture) are discussed (in relation to
e-Learning).
2.1 e-Learning
E-Learning, “is the use of Internet technologies to deliver a broad series of solutions that
increase knowledge and performance” [Rose01]. ROSENBERG promotes e-Learning as a
stand alone solution which doesn’t afford additional methods or methodologies to reach the
teaching (or learning) goals. HENRY defines e-Learning as an “appropriate application of
the Internet to support the delivery of learning, skills and knowledge in a holistic approach
not limited to any particular courses, technologies, or infrastructures” [Henr01]. Both use
distance education and e-Learning synonymously. GURI-ROSENBLIT claims that there are
huge differences between “distance education” and “e-Learning” at least because “distance
education in most higher education systems is not delivered through the new electronic
media, and vice versa – e-Learning in most universities and colleges all over the world is
not used for distance education purposes” [Guri05]. The Swiss “workgroup e-Learning” of
the SKPH (Schweizerische Konferenz der Rektorinnen und Rektoren der Pädagogischen
Hochschulen) recommends following definition as the official one (at least valuable for
Swiss): As e-Learning, every type of learning is meant, in which digital Media for the
transmission of learning material, the communication or to support the collaboration are
used (free translation from German) [SKPH07]. Since a generally valuable definition of e-
Learning doesn’t exist within the common literature, I adopt the approach of the Swiss
“workgroup e-Learning” as the one used in this paper.
This paper primarily focuses on course-modules and on the smallest possible entities
within e-Learning (i.e. single illustrations). Nevertheless, a full course can also be seen as a
module (i.e. of a bigger course). In our days, e-Learning developers strongly face the re-usability
of contents (i.e. [StCo06], [FEMA02]). The maximum reusability-rate can be
reached in a modular system, where even smallest parts of courses (like single illustrations
or chapters of a text document) are selectable through explicit descriptions (metadata) and
exchangeable. In the future, instead of writing new courses in its full, new didactically bal-anced
and semantically rounded courses will be constructed by selecting and combining
those modules. Self contained courses, which are written in one piece like self training-courses
stored on a CD-Rom surely still will be used in the future but it can be expected
that they will loose in their importance. The herein defined criteria should fit to classify
those kinds of courses, too.
10. 2.1.1 e-Learning Environment (Definition)
As the term e-Learning itself, e-Learning environment is not clearly defined, neither.
MÖDRITSCHER focuses on the software by claiming that “an e-learning environment has to
provide methods to adapt to the learner as well as to the teacher”. He adds that the “e-
Learning environment may adapt to the learner with regards to his characteristics” and
that it “can adapt towards the teaching process, which can be described for example with
the preferred teaching styles of a teacher.” [Mödr+04]
In this paper, the term “e-Learning environment” includes everything relevant to e-
Learning, which is susceptible by the rules which are defined for the system and by the ac-tors
within the system: The software which is used (i.e. Learning Management System –
LMS, course-contents), the hardware (technology) which is needed to make e Learning ac-cessible
and usable (i.e. PCs or communication technology), the chosen didactical ap-proach
(i.e. behaviourist, cognitivistic, constructivist) and instructional design, the organi-zation
(who has which duties and rights, how are processes implemented) and the social
environment (i.e. the role bearers themselves).
2.1.2 e-Learning Context (Definition)
The context of e-Learning is everything what influences e-Learning environments but can
not (willingly) be influenced by the environment itself. A further description shall be given
in the Chapters 2.3 and 4.
2.1.3 e-Learning Context-Metadata (Definition)
Metadata, mainly used in concerns of databases are defined as being data about data
[W3C98]. In this meaning, the context-metadata which are to be defined herein the docu-ment
describe the data-structures and already-represented contextual attributes, related to
certain fields of context and which play a role as an influence factor on e-Learning. Never-theless
they also are data about data, because the attributes themselves are describing short-forms
(like key-words) for larger coherences. In [IEEE02] metadata are defined as “infor-mation
about an object, be it physical or digital”. Corresponding to this and in the mean-ing
of this document, metadata are such data, which describe significant parts of a whole in
the form of defined keywords.
Context-metadata represent the influences on e-Learning courses which directly can not be
affected by the course authors, the environments or the courses themselves. They give
hints on the compatibility of modules and describe attributes which are to be considered
within the targeted region. Those attributes are to be considered during the adaptation pro-cess
(internationalization, combination of modules) to find out changing needs. Context-related
criteria not only are needed concerning general aspects of the e-Learning context,
but also in specialized contexts like company-environments and related to special subjects.
This increases the potential number of context-metadata enormously. In this paper,
primarily possible shapes of context are to be described to find a useful order-system and
metadata on a general level. Concrete attributes for data-structures will only be defined in
well describable or in generally applicable cases. In the following, a metadata contains at
least a metadata-name and one attribute, which can be an open definition (wild card).
3
11. 2.1.4 Adaptation of e-Learning (Definition)
In the common e-Learning related literature, “adaptation” mostly relates to pure technical
adaptation in special concerning adaptation of content for mobile technologies ([Ton06+],
[GoKi04], [GoKi06], [Vite00], [SpKr06], [Ble+05]. TONG et. al. write “Although content
adaptation techniques have been extensively studied for mobile computing systems in last
decades, most of the previous work focused on adaptation with respect to terminal capa-bilities”
[Ton+06]. HAN et. al. define content adaptation as a “process of selection, genera-tion
or modification of content (text, image, and animation, etc.) to suit to users’ comput-ing
environment and usage context” [Han98+]. Additionally they describe content adapta-tion
in the form that “it can be applied to transformation within media types, such as re-ducing
image size or resolution, and across media types, like converting speech to text, or
video item to image montage” [Han98+]. GRAF & LIST [GrLi05] differentiate between
adaptability and adaptivity in a way that adaptability “includes all facilities to customize
the platform for the educational institution’s needs (e.g. the language or the design)” and
“adaptivity indicates all kinds of automatic adaptation to the individual user’s needs (e.g.
personal annotations of learning objects or automatically adapted content)” [GrLi05].
BEKIARIDIS uses the term “localization” instead of adaptation and writes “Localization is
the process of adapting a product or service to a particular language, culture, and desired
local ‘look-and-feel’ [Beki03]. ROSMALEN et. al. reduce the adaptation process on the
learner’s needs: “Adaptation in the context of e-Learning is about creating a learner ex-perience
that purposely adjusts to various conditions (e.g. personal characteristics, peda-gogical
knowledge, the learner interactions, the outcome of the actual learning processes)
over a period of time with the intention of increasing pre-defined success criteria (e.g. ef-fectiveness
of e-learning: score, time, economical costs, user satisfaction)” [Ros+04]. A lot
of researchers, especially such working in the educational context claim that only small at-tempts
have been made to do research concerning context-aware adaptation processes (i.e.
[Sta+04], [Wolf03], [Spi+02]).
In this paper, the term “adaptation of e-Learning content” includes all processes which are
needed to make e-Learning content (representing the information, which is to be taught)
written for an originating context accessible for another context. This includes all neces-sary
changing work i.e. concerning didactical strategies and methods. The result of the ad-aptation
process shall be a course implemented in the targeted society which’s output is
adequate to the one of the originating country, where it has been written or at least differ-ences
4
in the output are to be known.
2.2 Standards - a Brief Overview
In this chapter, the common e-Learning standards which contain descriptions of metadata
shall be analyzed, if concrete context-related metadata already are defined. If they already
are defined in a sufficient detailed way, they do not need to be defined (but mentioned)
again. In addition, the special structure (terms) how metadata are presented in the common
standards, shall be analysed on its usability for the context-metadata, which are to be de-fined
in this paper. Possible adaptation-needs concerning the way metadata are presented
(attributes …) shall be pointed out. If the presentation forms used in the common specifica-tions
are not useful, maybe alternative presentation forms can be developed to provide
maximum compatibility to existing standards. The e-Learning related standards and speci-fications
firstly are made to provide interoperability between related applications and con-tents.
The term “quality” in e-Learning until now mostly is considered procedural and does
not meet the content itself. Not all specifications deal with definitions of metadata. Those
which do so, define own (also specialized) sets of metadata, define metadata structures,
12. conclude other metadata specifications to harmonize them and/or use already defined
metadata-structures and metadata for procedural descriptions.
The related and herein discussed specifications are Learning Objects Metadata (LOM)
[IEEE02], which in special deals with personalization, course-content and course require-ments,
Learning Design [IMS03], defining didactical aspects, DIN-PAS 1032-1 [DIN04a]
and DIN-PAS 1032-2 [DIN04b]. The DIN-PAS 1032-1 defines basic constraints of e-
Learning courses and 1032-2 historically seen was an addition in which some context re-lated
metadata are defined. Context-metadata in their impact are far more than only a side
aspect of boundary conditions: the DIN-PAS 1032-2 implicates this through its position as
annex. This “addendum-philosophy” used for the DIN-PAS 1032-2 suits with SPECHT’S &
KRAVCIK’S demand: “The current metadata sets should be extended for capturing and
handling additional context data” [SpKr06], but I go some further and recommend a defi-nition
of context-metadata as a stand-alone specification to point on the upper described
impact-depth and prevent possible evaluations (as “it’s only an addendum”). [DINI05]
considers Metadata for “electronic teaching- and learning materials” as well as metadata
for multimedia objects. The Dublin Core Standard [DCMI06] which relates (a part of the
standard) to structural issues concerning learning-metadata has to be analysed. “Metadata
for Learning Resources” [ISO05b], [ISO06] as a representative for specifications which
define metadata-sets for special demands by considering existing specifications with the
purpose to simplify the definition processes for users, shall be analysed, too. There are fur-ther
specifications which pick out certain aspects of other metadata definition specifica-tions
for special purposes, like Public and Private Information (PAPI), which only con-cerns
Lerner and Teacher Information [IEEE01] and Computer Managed Instruction (CMI)
[IEEE04] which primarily focuses on data-modelling using the AICC standard and the best
practice guide CanCore [Canc06] (also see 2.2.2).
A special position between the standards does the Aviation Industry Metadata Description
of the Aviation Industry CBC (Computer Based Training) Committee (AICC) have: It’s a
standard defined in and related to a certain business (Aviation). The AICC coordinates its
efforts with other standardization groups, like ADL, IMS and IEEE LTSC. The metadata
standard will be also analyzed in here, because companies provide their own contextual
characteristics within the field of e-Learning (see also 2.2.8). Another specialized standard
(here on business processes) has been defined by the ebXML group of the UN/CEFACT
and OASIS [UN01a], [UN01b]. Within this paper, metadata, representing legal aspects are
specified for the business process. The paper itself will not be discussed in detail in this
section, because there is no direct relation to e-Learning. But briefly within the Chapter
5.5, which deals with law-related metadata, it will be taken into consideration.
Structural presentation definitions for metadata are to be proved if they are usable in the
context-field, too: Adapting those would provide a basic compatibility to already existing
standards and as already compatible a better acceptability within the community. Finally,
SCORM [ADL04] has to be mentioned and briefly described (2.2.9), at least for comple-tion.
SCORM is the standard of the Advanced Distributed Learning (ADL) group and pri-marily
has been made to provide (e-Learning) interoperability (content-related) between
the different sections in the American Department of Defense. The standard doesn’t define
metadata structures nor metadata but uses (considers) the standards DC and LOM.
2.2.1 Learning Design (LD)
The Learning Design (LD) specification [IMS03] of the IMS (Instructional Management
System) Global Learning Consortium (IMS) is activity based and focuses on online learn-ing
situations as well as offline learning situations (i.e. CD-Rom based courses). It imple-ments
three different modelling levels (A, B, C). While A is purely conceptual, level B and
5
13. C are more detailed. It focuses on the learning process (input/output), the instructional and
didactical methods as far as they are directly involved into the learning process, and on
persons (learner and teaching/tutoring staff). Nevertheless some side aspects mentioned in
the description of the more general “Level A” are included in the specification, which may
be interesting to be taken into consideration although they are used content related in this
specification: The environmental description includes the possibility to define “Generic
Services” (p. 12). “Generic Services” allow the definition of communication applications
and needed hardware/software to make the content in the form available, in which it is re-quired.
A single digital pencil needed by the student to work with a virtual workbench (in-cluded
in a learning situation) would, for example, be defined as an “item model” corres-ponding
to the data-model. All necessary technologies needed for a single course are to be
defined (list) in the Information-Table “environments”. This list contains more information
than only the technical environment (also learning objects), so it could be difficult to use
this list for a comparison. An adaptation (i.e. separating the list into an item-related envi-ronment
list and a content-related list) of this list to make it accessible for a context-metadata
specification in the future could be useful. Concerning communication technolo-gies,
additionally the Information Table “service” defines e-Mail and Conference-manager
services. Those at least are LMS- and learner-focused and because of the data-overhead not
usable for context-metadata. The technology related data within LD surely are focused on
the required technology concerning the availability of the course (learners must have) but
the same format could be used for related context-metadata, which belong to countries,
companies or regions. If this data-set already is defined for a course as requirement, it may
be a simple task to compare it with the data-set for expectable hardware and communica-tion
technologies within the (for adaptation) targeted environment. Also the “support ac-tivity”
which is defined in could be structural reused to define expectable skills concerning
the tutors which are focused in a targeted country, region or company.
Concerning the stakeholders, the LD is focused on personalization and as explicitly writ-ten,
it is not possible to make general statements (instead of individual ones).
2.2.2 Learning Objects and Learning Objects Metadata (LOM)
Learning Objects (LOs) are not generally defined yet and so there is a discussion about the
being and use of them within the community. The metaphor used to describe Learning Ob-jects
as a kind of LEGO-stones, which freely can be combined is a simplification, which
doesn’t reflect the reality as WILEY expresses in his article “The Post-LEGO Learning Ob-ject”.
[Wile99]. If Learning objects could be compared with LEGO-stones, this at least
would mean that they have to be designed “context free and of small granularity”
[Gobe05]. GOBÉE adds that such requirements on LOs “would be troublesome when trying
to develop LO’s in practice, as they oppose with pedagogical ideals, or demand much ef-forts
without direct returns”. As alternative WILEY in [Wile99] proposes the metaphor of
an atom, because not every atom is freely combinable with others and certain preconditions
have to be fit. Also, the metaphor of an atom doesn’t imply that combining LOs is as sim-ple
as combining LEGO-pieces. Both, GOBÉE and WILEY, are of the opinion that “Learn-ing
Objects cannot be free of context” [Gobe05]. WERTSCH says that “action is mediated
and cannot be separated from the milieu in which it is carried out” [Wert91]. Free from
the debate of metaphors, FERNANDEZ-MANJON & SANCHO try to define LOs: “The idea be-hind
learning objects is clearly grounded in the object-oriented paradigm: independed
pieces of instruction that may be reused in multiple learning contexts and that fulfil the
principles of encapsulation, abstraction and inheritance.” [FeSa02] The Institute of Elec-trical
and Electronics Engineer’s (IEEE) [IEEE02] defines LOs as “any entity, digital or
non-digital, that may be used for learning, education or training”.
6
14. Learning Objects have to be referenced by metadata. The draft standard “Learning Object
Metadata” (LOM), which is a cooperation project between ARIADNE, IMS and the IEEE
Learning Technology Standards Committee (IEEE LTSC) “defines the structure of a
metadata instance for a learning object”. Considering IEEE LTSC, Learning Objects in
this meaning can be i.e. multimedia content, instructional content, learning objectives, in-structional
software and tools or persons, organizations or events. The idea behind LOM is
to make the explicit identification of Learning Objects possible by describing specific at-tributes,
what is a crucial condition for platform independent processes concerning search-ing,
finding, division and reuse. “Typical attributes include the type of the learning object,
author, owner, terms of distribution and format of the learning object. Additional support
for pedagogical attributes is provided, like interaction style, grade level, mastery level or
prerequisites” [VePa05]. LOM at least is the harmonizing attempt to conclude the stan-dardization
efforts of the IMS [IMS03] and ARIADNE [ARIA02] (ARIADNE is an appli-cation
profile). In LOM, a lot of metadata classes and types are defined and the biggest
subject of critics is the resulting lack of manageability. As consequence of those critics, the
Canadian approach of a simplification of LOM, the specification “CanCore” (Canadian
Core Learning Resource Metadata Application Profile) has been developed. It is seen as an
“instantiation of the LOM-standard” to “facilitate the interchange of records describing
educational resources and the discovery of these resources” [Canc06]. “CanCore is based
on and fully compatible with the IEEE Learning Object Metadata standard and the IMS
Learning Resource Meta-data specification” [Canc06]. Because of the already reached
complexity of LOM, it is recommendable not to define the context-metadata within LOM
but (as already stated before) to define a separate specification. Additionally, the coexis-tence
of various standards besides LOM lead “to a lack of interoperability between de-scriptions
based on different standards and approaches [ISO05b].” As an alternative to
LOM, the specification “Metadata for Learning Resources” [ISO05b] has been defined
(working draft). It will briefly be discussed in 2.2.3.
Since context can not be described like, nor interpreted as Learning Objects, it is not useful
to prove in detail if the context corresponds to already in LOM defined LO-categories. Dif-ferent
contexts and related subjects may have very different attributes to define, because
they cause very different learning situations. Because of its complexity, basing on the vari-ability
and dependencies from each other (as will be shown later) context can not be seen
as an object but far more as a reason for a certain situation within an environment.
The LOM-standard defines 9 different categories which are represented by each one data-set.
Besides analyzing the predefined datasets the general adaptability for and possible
lacks concerning the representation of context related metadata shall be shown. Every data-set
in LOM consists of a set of categories, each represented as a table-row consisting of the
7
following attributes:
| Nr | Name | Explanation | Size | Order | Value space | Datatype | Example |
The predefined datasets aren’t meant to describe context but Learning Objects. Necessary
fields to describe contextual dependencies aren’t included. It is not possible to link table-rows
with each other. Direct connections between table-rows are not supported as it would
be needed in some cases of contextual metadata to reserve the intuitive consistency of data
and avoid redundancies. Hierarchical structuring concerning the dependencies of data-sets
is not supported but this could be tricked with the field structure (General, 1.7). Neverthe-less,
even when this would be possible, only a single hierarchical structure definition for
each learning object is defined but not a hierarchical structure for each attribute of Learn-
15. ing Objects. At least it is questionable if tricking a system is useful when machine-aided
interpretation is targeted. Arranging data within an intuitive ordering system is crucial (as
it will be shown later on) to provide the necessary manageability to the authors, who later
on shall define the metadata for their courses. The attribute “Order” only shows if the de-fined
attributes (if there is more than a single one) within one row of the table are ordered
or not and is not useful for this purpose, neither.
A concrete case where dependencies as well as hierarchical structures have to be shown is
the language: the language tag in LOM (“General”) can only be used to define a concrete
used language (or more than one), but not to directly point at the cultural background of
the author, who wrote a course within this language or to point on a certain used regional
interpretation scheme (or “sublanguage”). There is a difference if two languages can be de-fined
simultaneously or if a hierarchical order (i.e. country language, social dialect) can be
defined: special adaptation needs can only be found out when the necessary information is
available.
It is possible to show the version number of a course (through Life-Cycle in 2.1), but not if
the course already has been translated in another language or adopted into another context
(i.e. from school to company) or in special, which changes have already been done. The
originated language may have changed to another language (i.e. because both countries are
Western European ones) without having adopted the specific cultural attributes of the
author: The language then is not an indicator for culture anymore.
The dataset “Identifier” in “Meta-Metadata” (3) generally could be used for context-metadata,
too. If it’s meaningful is questionable, because in special here the possibility to
reflect hierarchical dependencies is missing. The predefined dataset for “Technical” as-pects
(4) could be reused for context, too. This has to be checked (and discussed in Chapter
6) in special cases, but most necessary data seem to be taken into consideration to fit for
technical descriptions for targeted context, too. The “Educational” dataset (5) only takes
the concrete content into consideration. In the given form it can not be used to describe
educational issues in general or even certain context although in 5.1 there is a metadata
“educational context” included. It is directly related to the course (content) but not usable
apart of the content. The dataset “Rights” (6) only considers learning resources and with it
intellectual property rights and conditions to use the resource. Needed in this concern for
the context, are definitions of various laws (see also Chapter 0) which are to be taken into
consideration while adapting a course. This table could easily be expanded in LOM, so that
compatibility between the datasets of the course and the targeted context is given by re-flecting
the corresponding laws ruling in the originator’s country. The Dataset “Relation”
(7) is content related and describes relationships between Learning Objects. This dataset is
not usable, because there is not a pendant in the contextual environment. The same can be
said to the dataset (8) “Annotation”. The Dataset “Classification” (9) also is related to
Learning Objects.
2.2.3 Metadata for Learning Resources (MLR)
This ISO-document, which is not a new standard (yet) but a recommendation paper is covered by
two papers: the [ISO05b] and Information technology - Learning, education, and training -
Metadata for learning resources - Part 1: Framework [ISO06] and the Working Draft for
ISO/IEC 19788-2 – Metadata for Learning Resources - Part 2: Data Elements [ISO05b]. It
has been designed “in order to both achieve interoperability and consistency with IEEE
LOM 1482.12.1 and other approaches” by using a “description of a conceptual level inde-pendently
of any particular representation” [ISO05b]. In this papers, “learning resources”
are focused and not “Learning Objects” which on the one hand shall free from the discus-sion
what LOs particularly are and on the other hand allow those standards which do not
8
16. define LO’s (like Dublin Core, see also 1.3.6) to adopt the predefined resources, too. Addi-tionally
neither a specific environment nor a description of such is forced, to provide the
maximum possible compatibility. Founding on the conditions the IEE/IEC 11179 standard
(general Metadata description standard), 6 core elements have been defined, which all con-tain
further sub-elements. Those sub-elements do not need to but can be defined, what
makes the definition process much more easily manageable than using i.e. LOM. Addi-tionally,
the core elements themselves can be used as a container structure, so that they al-low
the free definition of more specific sub-elements. The core elements are Description
(intellectual content of a resource), Instantiation (information to identify one or more in-stances
of the resource), Contribution (Information about contributions to the resource),
Contextualization (information about the environment in which the resource is intended
to be used), Access (usage conditions) and Record (Registering information, description of
the record itself). Adopting the whole specification is not possible, because the option not
to define course related attributes (they are mandatory to be defined) is not given. The
documents still are in working process and it can be expected that the content will change.
An evaluation on the usefulness of MLR for context-metadata follows in 2.2.10.
2.2.4 DINI / ELAN: Metadaten für elektronische Lehr- und Lernmaterialien
The working group “Metadaten für Multimediaobjekte” of the initiative DINI AG
(Deutsche Initiative für Netzwerkinformation e.V.) together with the working group
“Metadaten” of the „eLearning Academic Network Niedersachsen“ (ELAN) defined in the
Document ELAN AG “Metadaten” [DINI05] metadata specification. Targeted are the con-tent
(“Lehr- und Lernmaterial”) and the actors in the e-Learning scenario (Universi-ties/
schools, teacher and students). Other standards like Dublin Core (see also 2.2.5) and
LOM (see also 2.2.2) have been taken into consideration to provide maximum compatibil-ity.
As in MLR a core-set of mandatory to define metadata has been developed in special
to provide the best possible interoperability between different learning management sys-tem(
s) (LMS). Also, optional to define metadata are provided. Those (both, mandatory and
optional ones) are summarized in the “ELAN Application Profile”. In their specification,
they differentiate between content and course: In their definition, a single course may con-sist
of various contents. The specification primarily represents functions, which are typical
for libraries: It is focused on concrete course related subjects like who may teach a course,
what do learners have to know as precondition and what the content of a course is like, the
specification is not transferable on the pure context of e-learning in a useful way. It shall
not further be taken into consideration within this paper.
2.2.5 Dublin Core Metadata Schema
The “Dublin Core Metadata Terms” [DCMI06] of the Dublin Core Metadata Initiative
(DCMI) is a general Metadata specification supporting a “broad range of purposes and
business models” [DCMI website]. Besides other subjects “DCMI’s activities include …
educational efforts to promote a widespread acceptance of metadata standards and prac-tices”
[DCMI website]. The DCMI metadata are all similarly structured (but different to
the structure in LOM) and the predefined educational purposed related metadata concern
the content or the runtime of certain courses. Because of the large number of mandatory to
be defined, course related attributes this standard can not be used to define a context with-out
considering a concrete course. Although DCMI conformed metadata can be self de-fined,
it seems not to be useful to adapt this specific standard because it is not combinable
9
with the LOM-standard.
17. 2.2.6 DIN-PAS 1032-1
The specification DIN PAS 1032-1 (PAS – Public Available Specification), which is de-fined
by the „Deutsches Institut für Normung e.V“ (DIN), is meant to provide a useful re-ference-
model for quality assurance [DIN04a]. The additional purpose is to define special
didactic demands concerning e-Learning. That for processes for planning, development,
execution and the evaluation of education-processes and -offers (in special e-Learning) are
identified and described. Part of those process descriptions are descriptions of the targeted
(with a course) internal and external context considering social and economy related
boarder conditions (see also 6.1.6, 6.10) and boarder conditions related to laws (see also
6.5). Concerning companies (see also 6.13, 4.1), actors (and their expected experience /
knowledge – see also 0), business models and the organisational structure can be defined,
as well as the internal learning culture. Those data later on in Chapter 8.2 have to be dis-cussed
for their use, concerning the also later on defined comparison process in Chapter 0.
Further more concrete criteria are defined within the specification, which can be used to
evaluate the quality of e-Learning products. The DIN-PAS-1032-1 doesn’t define metadata
in special, so for this purpose it doesn’t further need to be taken into consideration within
this paper.
2.2.7 DIN-PAS 1032-2
The purpose of this additional document [DIN04b] is the same as in DIN-PAS 1032-1. In
1032-2 metadata are defined along a defined learning scenario. In the focus, stand DOs
(Didactical Objects). In contrary to other specifications the DIN-PAS 1932.2 takes consid-eration
of the context of e-Learning as part of the application-environments of a DO. De-scribing
the context of a LO, the four attributes “Name”, “ID”, “kind” and “type” are to be
defined. A weakness of the specification concerning the description of context-metadata is
that the metadata are directly bound on a DO. A further evaluation of this specification fol-lows
10
in chapter 2.2.10.
2.2.8 AICC – Aviation Industry Metadata Specification (AIMS)
The Aviation Industry Metadata Specification [AICC06] focuses Learning Objects as said
in the introduction by “A metadata system is a set of parameters associated with a learning
object which enables a potential user to search for and evaluate that object. This informa-tion
can also be used to track and organize the production process of the training courses”
[AICC06]. Since AICC is actively participating in the work on the LOM standard, the in-dustry
metadata standard (by now) not only uses the systematic of LOM but also all there
defined metadata. “All categories and data elements added by the AICC appear after the
LOM categories and elements” [AICC06]. The specification has the purpose to implement
a specialized data set for aviation related learning activities. The context-field from LOM
has not been adapted. Instead, an own context related field has been defined in AIMS un-der
4.5.16, called “instructional context”. This context field is also connected to a concrete
learning resource and provides 5 different shapes, which all describe the context a Learn-ing
Object is meant to be implemented in (simulation, performance_support, on-the-job,
classroom, individual study, actual_equipment). This specification doesn’t fit the require-ments
to define the context of environments apart from concrete courses.
18. 2.2.9 Sharable Content Object Reference Model (SCORM)
The ADL-Standard SCORM [ADL04] is not a standard in which metadata are defined, but
a standard for learn management systems and learning resources, considering the metadata
(definition rules), defined in LOM and DC. Its purpose is to realize a simple exchangeab-ility
and a better reusability of contents within different environments and a standardized
interface to access them. SCORM consists of 3 functional packages with different respon-sibilities:
First of all, a Run-Time-Environment (RTE) is defined, which is responsible for
the sequencing of learning content and the user administration in runtime. The Content
Aggregation Model (CAM) describes all the resources which can be used in relation to
Learning Modules as well as opportunities to restructure and combine contents to be dis-tributable
(i.e. into other environments). Such restructured contents can be ordered within
“organizations” (hierarchical models). The CAM uses metadata as they are defined in DC
and LOM. The last functional package is available since the latest version of SCORM
[ADL04] has been released, but unfortunately, it is not yet supported by most of the al-ready
before release date implemented authoring- and learn-management-systems. It is
called the “Sequencing and Navigation (SN) specification” and is responsible for the vari-ation-
possibilities which are given for a user to navigate through a course. In special con-cerning
learn management systems and course development (besides content), SCORM as
a standard has a high relevance in our times.
2.2.10 Intermediary Conclusions
A problem which gets clear after having analyzed, the upper standards is that they all are
written in relation to and directly bound on courses. There are two different ways to look at
context: out of the view of an existing course which belongs to the context and in which it
has been written or is taught and out of the context of a possible environment which is not
related to the course but shall be a location where content maybe is going to be imple-mented.
In addition, all the currently existing specifications mandatory require the defini-tion
of metadata concerning concrete courses, as the actors, the course’s subject, course re-quirements,
a.o. – a situation which is not given.
If a course shall be adapted to a different context, we also need metadata describing this
targeted context. What we need are metadata which are flexible enough for describing very
different subjects (as it will be shown in Chapter 0) apart of course descriptions. The best
fitting solution would be (additionally) attaching the corresponding (newly defined) con-text
data-set(s), to every (new) content, which is to be written. In this way, the already used
specifications for content do not need to be changed (there only is an additional set of
metadata attached to every course), but it would be possible to compare the originating
context (freshly attached) of courses with any possible context (newly defined), in which
the courses could be adapted. In the following, some comments to the upper specifications
shall be given:
MLR: In special, the category “Contextualization” shall later on be proved if, in which
concerns and how far it can be used to similarly describe context related metadata apart
from describing concrete learning environments.
DIN-PAS 1032-2: The four available definition categories concerning context (Name, ID,
kind and type) do not meet the full requirements for contextual metadata but are so essen-tial
that they will have to be defined within the context-metadata classification model. The
context-part of this specification could fully be adapted (but has to be enhanced): The field
“type” can be used to describe consequences caused by a certain value of a concrete con-
11
19. textual attribute. Name and ID are purely administrative fields (a written name and a
unique identifier) and should only be used as such. The field “kind” allows to define an or-dering
system. With this field a hierarchical structure could be implemented. What is
strongly missing within the DIN-PAS 1032-2 is the support of directly or indirectly linking
to related contextual metadata and of an inheritance system including the definition of ex-ceptions,
which at least allows the consequent definition of substructures (“kind” doesn’t
fit). A special kind of society may have a typical behavior (i.e. a high power distance – see
Chapter 5.1.2), which only has to be defined once for all similar societies. An attribute
high context society (see also Chapter 5.1.4) has certain consequences for requirements
concerning course language and teaching style. The impact of the consequences addition-ally
is different with the level of the attribute and it also depends on combinations with
other attributes and their special attribute-value. So being able to link various metadata is
crucial to portrait this. Also it is needed to determine certain impacts in relation to the at-tribute
values (see also Chapter 4.2). An additional weakness of this specification as al-ready
mentioned in Chapter 2.2.7 is its position as annex to the DIN-PAS 1032-1, which is
concerned with didactical demands and quality insurance issues (see Chapter 2.2.6).
Someone who needs a specification for contextual metadata would not search it in this
place. Adapting the metadata structure may be partly useful at least in a way that the new
context-metadata are compatible to the context-metadata description by using the 4 rec-ommended
description fields. Besides the direct binding of the existing specifications on
courses, additional fields will be needed to describe the context: Thus a completely new
specification is recommended.
2.3 The Context of e-Learning (Brief Description)
In the following, a brief overview shall be given concerning the structures of the subchap-ters
of Chapter 0 by showing basically different influence factors. The different classes of
influence factors concerning e-Learning are represented by each subchapter. Since culture,
as will be shown in Chapter 5 could be seen as something similar to a super-class for con-text-
metadata of learning environments [Gann04], [CheMa98], [Rog+06], because it influ-ences
a lot of other influence factors, too [LeCo06], [CaEi83], [McOl00]), I start with ana-lysing
definitions of culture [Maba03], [KiNk05], [Dei00], [Coll99] and prominent models
how to describe and compare cultures [HoHo05], [TrHa06], [HaHa90], [Hend96]. All
models directly are also analysed concerning their impacts and usability in e-Learning.
Afterwards in the document, more specific culture-related literature concerning influence
factors will be analysed. Concerning the metadata-class “culture”, I found some basically
different (to each other) sub-classes during the research which will be deeply discussed and
analysed later in Chapter 0:
• Language [Leon02], [DaJo02], [Davi06], [DeMa06]
• Ways of communication [Cak+02], [Ting88]
• Humor [KiNk05], [Kend06], [KiNk05], [LiLe07]
• Acceptance of media types [Will02]
• Gender differences [Hase00], [Hen07], [BeSt02], [Simo04], [MaHe06], [Cock99]
• Social capital [Mer+02], [Stone01], [Liu07], [PaSl01], [Mer+02]
• Habits and preferences [Noc02], [DuMa07]
• Acceptance of technology [HeNk06]
• General opinion [Bea+06]
• Taste [Bour92], [Loh+06], [Gans99]
• Indigenous cultures [UNES05], [Dys+06], [HuDa07], [Auld02]
• Hints for technology transfer [LaDy06], [LeSm94], [Dyso03], [Henson90]
12
20. • Pedagogical approach [Mcca07], [DuMa07]
Besides this single metadata-class “culture”, the order in which the following chapters are
structured to each other is random. As it will be shown later, the impact of the single influ-ence-
factors can be very different depending of their intensity and combinations with other
influence factors. Trying to absolutely evaluate the single classes in relation to their impact
on e-Learning and each other seems not to be useful. Literature to additional (to culture)
context-classes, which allowed extracting potential metadata (and not necessarily had to be
applied to culture), has been found in the fields of:
• Demographical development [BaBo97], [UN06], [ChWe04], [ChBa99], a.o.
• Religion [Pitt06], [Asch06], [Pau80]
• Technical infrastructure [Edmu07], [Seli04], [Heat01], [RsF06], [GunDK05], a.o.
• Rights [Lean05], [Marc04], [Kunz95]
• History
• Politics [Kear90], [Davi06], [Marc04]
• State of development
• Media richness [Wils02], [Gul+06]
• Financial aspects [GunDK05]
• Human actors [Zimm01], [Davi89], [Nielsen93], [Ada+92], [GoYu04], a.o.
• Rules: [WoKi04]
• Companies [Daga04], [DaSe04], [DIN04a]
• Geography and educational infrastructure [SaFo02], [LaZh03], [JaGr99]
• Learner satisfaction – Known demands [John00], [DaJo02], [TiLa04]
• Internet security [Kno+03]
Although the presented collection of context related influence-groups seems to be com-plete
at this point of time, it is possible that in future additional context-classes (see also
“context-blocks” in Chapter 3, page 14) will have to be defined. As it will be shown in
Chapter 4, the context of e-Learning is changing with the time and new kinds of aspects
can appear. Additionally the research founded on literature, which mainly has been written
in or translated to the English or German language. I can’t exclude the possibility that cer-tain
countries, cultures or societies provide unique aspects, which justify the definition of
an own context-class. This has to be considered when the specification for context-metadata
is written, so that the opportunity to realize such an expansion is given.
13
21. 3 Methods to Determine Metadata
In the following, (since it has yet to be evaluated if they all are needed) context-metadata
candidates are to be found on the basis of different categories. Most of the found categories
and data are directly or indirectly implied by the literature concerning intercultural (and
political, religious, historical, a.o.) concerns or in there described problems and regional
characteristics. Some resulted by further reflecting on the content of the literature.
A single country can contain very different societies with each different world views and
typical (learning) behaviours. It is crucial always to be aware of the fact that adaptation
work for e-Learning while targeting a certain country not necessarily reaches all inhabit-ants
(in special potential learners): A lot of countries in Africa are well known examples
for such: At the moment in Africa there are 13 known serious conflicts and 2 open declared
wars between societies within (not between) the countries [HIIC06]. There are about 118
known serious conflicts worldwide (those where brute force reins) and in addition, 160 yet
peaceful conflicts, which are to be solved but which also could escalate. In some cases, the
reason may “only” be a political one, but in other cases there are huge cultural differences
(i.e. Kurds, Sunnites and Shiites in Iraq).
The consequence for e-Learning and in special for the description of context is that a most
possible granular differentiation has to be realizable. In every single case for adaptation
work it (at least once until it is monitored) will have to be verified if there i.e. is a homoge-neous
culture, law, a.o. within the country. If there are inhomogeneous circumstances
within the country, it has to be found out if those differences although are compatible con-cerning
learning environments and at least, which region within a country in special is to
be targeted. To complicate the situation a bit more, it is not said, that those different cul-tures
are geographically separated. The Kurds in Iraq i.e. mainly live in the north of Iraq
and it may be possible to address a course specially adapted for Kurds directly to them but
it may not be so simple, to regionally divide Sunnites and Shiites within Iraq or other con-flicting
or just basely different groups of people within other targeted countries [HIIC06].
The problem of separation targets e-Learning in the moment when the learning scenario
includes phases in which the learners shall come physically together, i.e. for examinations.
Annotation: Because of the obvious necessity to clearly differentiate in concrete situations,
in the following the term “region” will mostly be used instead of “country” and includes
the possible reduction on a society, too. Nevertheless in most of the cases whole countries
are meant. Reversed, when the terms country or countries are used, it has to be aware, that
in special environments a region or a single society within a country may provide an own
context which, if targeted makes further adaptation of e-Learning necessary.
After having found an adequate (sufficient granular) set of metadata candidates at first, it
has to be discussed, if all of those may be necessary (maybe some are kind of too special or
at least unnecessary to be defined in general). Metadata sets should be as complete but also
small as necessary. Also necessary to be discussed is if it can be expected, that more con-text-
metadata of a certain type will be found later on. This would i.e. result in the defini-tion
of only named (but for now) content-less data objects. In the next step, a classification
system is to be defined which will be a reference on which the later data structures for con-text-
14
metadata will anchor.
On the way to the classification system some of the candidates for context-metadata may
appear repeatedly, since various contextual aspects implicate their needs. Groups or con-text-
blocks in the following are understood as accumulations of context-metadata which
22. belong to a same topic, i.e. culture related context-metadata. This overlapping of metadata
groups through their elements gives a hint, which groups stand in direct relation to each
other. The knowledge of those intersections is helpful as a preparation to avoid unneces-sarily
redundant data in a later on to be designed database model, which can be used to
store the context-metadata (not part of this paper). Although it complicates the ordering
system at first glance, overlapping will not be fully avoidable: On the one hand, the data
have to be modelled in a way, in which different constellations of context-metadata still are
possible (maximum differentiation ability) and on the other hand, the number of the groups
should be as small as possible, for that it is possible to intuitively handle them. Such a sys-tem
will not consistently be used in practice, if the system is unclear or if it is too hard to
deal with. Making a decision i.e. if content restrictions by law belong to a political or to a
juridical group would cause an unnecessarily complicated search for the author to define
the course correctly (if the decision can not be automatically realised). The ideal solution
would be a possibility to automatically search for the necessary data for each context and
add them to the course or use them to gain knowledge on targeted regions.
15
23. 4 The Context of e-Learning
In the following paragraph, first (4.1) the objects of the context itself are roughly named
and described. Their influence on e-Learning environments and in special their time-related
dependency to each other, are to be analysed and shown within a specially devel-oped
model (4.2.1). In Chapter 4.2.3 which parts of the context have to be represented by
context-metadata will be discussed.
The context of a learning environment includes all the factors that have an influence on it
without themselves being influenced by the learning environment. An author who writes a
course or a course module (in the following, both may be called “a course”) will address
this course to a defined group of learners in a known learning environment (i.e. a school
form, a certain company). He will choose a didactical paradigm, a design and last but not
least the subject, which the course deals with. Things he will not plan but which affect his
course, too are those influences which are natural part of his environment and view of the
world. His view of the world is influenced i.e. by his religion, the culture he lives in, the
rights system of his country he is familiar with and many other factors. In this meaning, the
course itself will contain a lot of not knowingly included aspects which form his cultural
environment and rather help his learners (who do have the same or a similar natural back-ground)
to successfully finish the course than bothering their learning success. But what
happens when this course shall be implemented into a different environment, i.e. a foreign
country or a company with different culture? In the ideal case, nothing negative happens
what differs from the already known success. But it also could happen that targeted learn-ers
misinterpret the course, do not understand parts of it (or all) or in the worst case diplo-matic
implications occur, which destroy the existing relationship between companies, uni-versities
or even countries. To avoid such complications, it is crucial to know about the
possible differences between certain environments which form the context where e-
Learning content shall be adapted to. The first step is the awareness, which can be reached
by naming potential candidates for such conflicts.
As a solution, it may be desirable to write a course in an neutral way but, if possible this
would also mean to forbear the author from implementing a didactical approach and all
those aspects, which also are part of his personal formula for success – it would also mean
to forbear from every personal note and at least to pass on the personal success, which may
be a crucial part of the author’s motivation to write a “good” course. Finally, the solution
in common cannot be an abdication of every aspect which keeps a minimum potential for
conflicts.
Even the “neutrality approach”, if meaningfully implementable, would not solve all prob-lems.
If course modules could be written neutral and at least only contain the pure informa-tion
this still would not guarantee that the course is adaptable without further work:
Finally, the provided information itself could be an obstacle: An example for such a con-flict
based on pure information could be historical information within a course module:
One country may see the “hard facts” i.e. about a war dept on historical information in a
very different light than others.
Well-defined context-related metadata concerning the author and the target group of the
course as well as their intended cultural, technical, political (a.o.) environments far more is
a solution for this problem. To adapt a course into a different context (environment) which
partly stands in conflict to the originator’s one means to find out the conflict potential and
change the conflicting aspects within the course (as far as possible). Because of very basic
conflicts, this adaptation not always will be realizable. Additionally in some cases, conflict
potential could be unforeseeable. Since obstacles may differ from environment (i.e. re-
16
24. gions, countries) to environment, this adaptation work will be depending on the targeted
contexts very different, too.
The discussion which design-choice is the best one, may not lead to an end because a con-crete
choice how to design it, is also depended on the intended use and it may remain a
philosophical discussion. Deciding which design, in general, is the best fitting one is not a
part of this paper. The context-metadata themselves, which represent possible conflict pa-rameters,
will further on be handled independently from the design-question.
4.1 Influence Factors on e-Learning
The targeted form of e-Learning within this document is a projection of any kind of educa-tion
in presence form to an electronic form: Not everything may be possible or useful to be
adapted to e-Learning situations. Additionally the electronic advantages which e-Learning
provides as the use of multimedia and the possible independence of time (see also blended
learning [Vali02], [Ches02]) are included. In this meaning most (since usually there is no
direct contact to the teacher) of the factors, which influence learning in the classroom style
as well as collaborative learning, experimental learning, a.o. are influence factors on e-
Learning environments, too. Furthermore, there are additional context-related data, which
only influence e-Learning but do not exist in the classroom situation like laws concerning
the storage of personal data and technological infrastructures. E-Learning provides the pos-sibility
to be spread over the Internet. An e-Learning course in theory can be accessed
around the whole world – but surely there are gaps concerning the understanding, i.e. the
language. In the presence-education model, the expression to this different situation would
i.e. be to send a German teacher to China and let him teach the Chinese children in the
same language with the same methods and the same didactical approaches, he usually
would apply in Germany. It is obviously that this would not work in the case of the lan-guage
but a lot of other differences between the two regions additionally avoid the teaching
effort at least in the same way. Concerning problems in such situations, a lot of literature
(coming from pedagogies, mostly concerning the subject “German as Foreign Language”)
exists: It will be referenced within the next chapters together with concrete examples, when
certain context-metadata candidates are to be found.
Before the concrete context-metadata are to be discussed, the scenario contained within the
monitored model has to be described. In the following, the learning situation in presence
learning is to be described and directly compared with the e-Learning situation to deduce
context related impact factors, which will later on be represented by the metadata.
Referencing to the upper picture of the German teacher in China, it would be the question,
what he has to change in his teaching style, behaviour, knowledge, teaching contents a.s.o.
that the Chinese pupils are able to understand his content, so that he will be able to have
the same teaching success (or similar) in China than before in Germany. In the context of
e-Learning, the role of the present-learning-form teacher, who is the one with the know-ledge
about methods and content and who has the ability to submit the contents is played
by the author (and in some situations the SME). Additional (mandatory) to the taught con-tent,
a tutor can fulfill the teacher’s duties concerning activities being done in direct con-tact
to the learner (in present learning both roles are “played” by a single teacher).
A classical teacher usually supports a limited and manageable number of students within a
school, university or other institution. E-Learning content once produced, in theory can si-multaneously
be spread to an arbitrary number of learners which (synchronously) would
not be manageable by a single person. The country in which the teacher has studied or
17
25. learned to play his role in e-Learning corresponds to the country in which an author got his
knowledge background and expectably writes the (originating) course in (and for).
Illustration 1: e-Learning Scenario 1
In the illustration 1 which primarily (but transferable to the general learning situation) is
related to the e-Learning scenario, it is symbolised with country A (A stands for a specific
country). A more detailed description of illustration 1 follows in Chapter 4.2.1. The
Learner as actor within the scenario in the frontal present style (here, every style of learn-ing
is meant, in which teacher and learners are together in one room) is a recipient of in-formation
and a bit less active than the e-Learning learner: He is being taught as a service.
The role, he primarily has to play within the present teaching scenario is to listen to what
the Teacher communicates. Different to that, in the e-Learning situation, the learner (usu-ally)
at first has to choose and actively download the content (produced by an Author)
and then study actively in self-learning form (controlled and self motivated). If there is a
tutor involved, he usually plays his role in a later phase (i.e. examinations, experience
works, collaborative working, a.o.). For every single country, a set of rules like rights sys-tem,
political background, cultural background, and others, which have influence on the
learning scenario can be determined. In this and the following chapter and in special in the
illustrations 1-2 those rules (as influences) shall be referenced as Influence factors.
Since culture itself as dominant influence factor on learning-environments is dynamic
[Maba03], the context of learning and in special of e-Learning does basically change from
time to time. Dependent on certain events such changes can have very extreme impacts
which force following changes in teaching and taught content. During my research, I have
found out that those changes within the environment do not happen all at once. Far more
there is a time delay as well concerning the different influencing sources (learner, author,
country, region and society) as also within the single “context-blocks” (see Chapter 3, page
14) like rule-set, political situation, a.o. A change in the political system (i.e. a military
coup) may i.e. cause a change later on in law system and far later in the curriculum but all
single changes may influence the learning environment. As a consequence, this means that
if we want to specify the context of a possible e-Learning situation in which a course shall
18
26. be applied, we have to update the context-metadata before and corresponding to changes
which happened since the data have been defined or used the last time: The resulting ques-tion
is when and in which way context-changes may appear and if they are predictable.
This will briefly be discussed in the Chapter 4.2 on the example of the German-German
history. In special, this time-delay has not been described before within the e-Learning lit-erature
and so I have developed a model to describe it (illustration 1-3), which will be
shown and explained in the following chapter. The model is a simplified approach which
in later times and in a refined version possibly could be used to automatically simulate the
time delay and experience-based calculate possible impact types.
4.2 Contextual Changes and Dynamics
Tutor and author as simplification and corresponding to the teacher-model in the illustra-tions
of e-Learning scenarios (illustration 1-3) are modelled as a single “position”, al-though
this in practice hardly will be implemented in this way. As already stated, there can
be monitored a time delay in which the single context-blocks react on external influences,
like political changes, changes in the curriculum or others. Depended on the level of chan-ges
(how extreme they are), the full context of e-Learning or only parts of it will react on
the influences or maybe, too - nothing happens within the context. In the illustration 1, the
situation as described in Chapter 4.1 is modelled. The e-Learning course, learner,
teacher/author, (including tutor), Country and Influence factors (rules) are modelled. In the
further chapters, it will be shown that the influence factors modelled here can have differ-ent
forms and that there can exist additional ones (i.e. companies and societies as addi-tional
factor to countries). The model does not describe the contextual situation completely
but for demonstration purposes concerning the time-shifted dynamical behaviour of the
context, this simplified model is sufficiently detailed.
A country A with a State (0,0) which is impacted by the Influence factor Set 2(A) at a
point of time T(3) in which the influence x happened afterwards has the State (2, 1). The
first value symbolizes the adopted influence factor with its current set-number and the sec-ond
value represents the number of adapted influences in total (it is a counter which always
razes by one, when the first value changes). T(x) shows the number of significant impacts
which could be monitored. Every time an event happens, it raises by 1 (it is a counter). It is
not necessarily said that the influences force an impact on the system. As it will later on be
shown, it is possible that the country adepts (and all other “actors” shown in the illustration
similar) influences later than they happen or do not adapt them at all (but others may do in-stead),
so the (first factor within the) state does not necessarily correspond to the recent
value of the Set. The second related number within the State (0,0) represents the total
number of changes.
A course C written or changed in this country A with the state (x,y) is then shown as a
course C(Ax,z), with z representing the number of changes done at the course and x repre-senting
the state of A in the time the change happened (similar to State and Set). It is im-plied
that a teacher/author acts corresponding to the certain shape (concerning the influ-ence
factors) of his country, since education mostly has to be oriented on decisions, the
government has made (the time near reaction is idealistically seen, cause there are private
teachers, too). So the teacher/author automatically inherits the state of his country. It
should be clear that this implication does not fully catch the true situation.
A Learner living in a certain country A is to be expected, that he adopts the time-spirit of
the country A and follows its laws and ideas, but he also can react on impacts of the influ-ences
a country does not react on. So he can inherit (first number) the countries rule set as
well as the influences directly. The second number is a counter again.
19
27. In this model, a quite idealistic view has been implied: in particular it has been implied that
the society’s picture of the world changes in the same time period in which the country‘s
rules change (a further delay parameter is not implemented). This may only be true in spe-cial
fields and under special circumstances i.e. rights systems changes may be immediately
adopted. The opposite example where surely nothing would happen immediately is a slight
change at the political system which happens within a democratic country. When i.e. in the
USA, the republicans loose the election and the democrat’s party wins it, it is not to be ex-pected
that this changes take place immediately in every shape of the country. So it is to be
expected that, in reality, also here a delay in changing between event and impact will oc-cur.
To model this correctly an additional remembering function to keep an older status in
mind would be needed. Also, the learners will surely not change their behaviour or mindset
immediately (maybe never) when the country is impacted by such changing influence fac-tors.
Annotation: The learner may take the change at a moment in which the country already is
a step further … so my stack-implementation of events may not be suitable to model those
circumstances more detailed and a list-implementation would be more useful. The extreme
simplified version of the model in here is sufficient to show the time delay in which outer
influences impact the learning system.
4.2.1 Using the “Influences Time Delay Model” (ITDM): Concrete Example
What may happen, when because of a certain event a new set of influence-factor-changes
impact a learning environment? An extreme for such a change has happened with the re-union
of Germany in 1989-90. In special, the former GDR has been influenced by those
changes concerning learning traditions. True, there have also been influences in the FRG,
but those rarely may have influenced the Complete-German learning situation (apart from
some i.e. history or politics related contents): Those shall not be focused further on. All of
the following information within the scenario is fiction but realistically.
After such extreme changes the whole regional context can be related. Impacts on learning-situations
(including e-learning) are expectable: After the reunion, the Eastern German
(GDR) didactical (also concerning content) system as well as the curriculum completely
changed to that of Western Germany (FRG). Also the rights system changed and in schools
and work environments the expected language skills changed: Learning English language
in school was unusual in the GDR - instead Russian language has been taught.
The delay concerning the full adaptation of the influences sides the students was enormous
since they not only had to learn a new foreign language (also to understand i.e. technical
documents, written in German) but also to trust other people, work together, openly discuss
about politics and history, a.o.
After 17 years there still is a notable cultural difference between parts of the Eastern and
Western Germany, which exceeds the expectable (and in Germany common) regional dif-ferences
(like between Bavaria and Hamburg).
This example shows that even massive changes in the environment, do not necessarily im-mediately
impact all aspects of a country. Nevertheless, every single time when changes in
the environment can be realised, in theory, the impacts on learning environments have to
be checked, too.
It further has to be taken into consideration, that the mental changes sides the inhabitants
started long before the reunion has been realized: On the one hand, lacking the necessary
information because of the GDR’s restrictions concerning the information-divide, the East-ern
German people finally have not been able to have the same cultural development like
20
28. the Western German people. On the other hand, all the time through, they prepared them-selves
in hope for this special event. This, at least, influenced the time delay in the other di-rection,
so that the example is not fully representative. As an extreme situation which hap-pened
in the near past and which is well documented, at least it is ideal for a demonstra-tion.
What would have happened to an e-Learning course, if there would have already been such
one written in the GDR before 1989? If not having been reviewed, it would still be the
same, stored i.e. on a CD-Rom: Former existing documents have not been adapted because
of their political influences and rather old fashioned contents (and didactical methods).
In the coherence of this paper, it is interesting if and what had to be done with such a
course to make it usable for teaching within the new situation (in the united Germany). The
situation is similar to adapting a course to a very different environment (i.e. other country).
It seems being obviously that the old course may not fit into the new democratic system
and that there are fundamental adaptation needs.
Independent of the concrete subject of the course, there may be influences (certain shapes
of influence factors) from the former system which willingly or unconsciously have been
adapted to the original course by the author and to the addressed learning environment
through the former politics, the targeted culture (targeted, since the attempts of the gov-ernment
to implement a system-conform culture did fail), the rights system, the author’s
and society’s self-picture, the relationship between learner and teacher, and others.
Not every influence factor forces the same level of adaptation need in every course-subject.
A political change i.e. may force a political course dealing with recent politics quite fast to
be changed. But a course which teaches how to analyze music or art may not be affected
by a political change. Differentiations are strongly necessary: It can be expected that the
last decision if a course needs to be adapted after a deep social, political or how else moti-vated
impact, is depended on a human mind. Although there might be a lot of events and
reactions of the system which do not force changing needs, the possibility forces a check.
Illustration 2: e-Learning Scenario after environmental impacts, 5 years after reunion
This makes a targeted automated decision far more complex since concerning the events
and reactions, there is no regularity. The ITDM can be used as a decision support system:
Based on formerly made experiences, ITDM in theory can evaluate, if changes will happen
21
29. and visualize when and where needed adaptations are expectable. How far this in praxis is
possible will be discussed in Chapter 4.2.2.
In the following it will be shown step by step by using the ITDM, how the system may
have reacted at the side of the east of Germany during the reunification-process. First cer-tain
events are briefly described and afterwards the reaction of the system. In the table in
end the events and reactions are summed up.
When the wall broke down, the general law of western Germany automatically was im-plemented
to the former GDR regions as well as specific laws (i.e. criminal law). Not
every single part of those specific laws has been implemented (i.e. administrative law) be-cause
a lot of compromises had been needed for the transitional period. The situation in the
eastern parts had been far too different to implement all at once. At first, the necessary in-frastructure
(i.e. policemen, various administrative posts) had to be staffed and the political
structure (Federal Administrations) had to be implemented.
Concerning the model, in this phase we have the influence factor set 1 in the time (point of
time significant changes happened) T(2). The inhabitants represented by learners already
have adopted some of the expectable changes in the past (this was T(1)). The eastern part
of Germany adapted the changes in law and some of the new possibilities, the western cul-ture
offered - and reached state 1. It was the first (1) change the country made but within
T(2): state (1,1) T(2). The learners who already adopted changes at the point T(1) which
the former GDR-authority did not adopt, now adopted the new laws (second change) and
reached also state 1 in T(2) which means L(A1,2) - T(2). The course stayed unchanged
and has not been taught anymore.
After some time, the monitored situation was clearer and the necessary infrastructure had
been widely implemented. The government brought further laws to validation. Since this
influence-factors set 2 impacted the country A, the event time-changing stamp moves to
time T(3). The country A adopted the law with all consequences and because of this its
state value moves to state (2,2) T(3) (second change in 3rd timestamp). The inhabitants
(learner) have been a bit overburdened with all the changes around them and since the new
laws did not inflict their lives directly, they did not see a reason to adopt the latest law (i.e.
government-internal administrative laws) anyways. So nothing has changed with the in-habitants
– in the model, they stay in their recent shape. The same (nothing) happened with
the course, which still remains in the state it had in times of GDR.
Next, T(4), a briefing from the EU may have arrived all the EU-States, stating that all
school contents had to be brought up to a certain state of the art. The monitored Eastern
German course was not taught anymore so the influences have been noticed but did not
force further reactions sides the government. The influence-factor set for now still stays 3
and the country does the 3rd change to synchronize the standard with the new law (there
was no significant impact on the country because it did not relate the GDR-course to the
law). The state now changes to state (3,3) in the change period T(4).
After some time the German government found out that the content of the old GDR-course
had a higher worth than expected. The course now had to be “reanimated”. The new Euro-pean
law for quality in learning material had to be achieved concerning the course.
In this case, there has no change happened within the country itself or the learner nor the
set but the course now after adaptation, represents the state of the countries art (“united-
Germany”): The state of the course jumps to the recent countries state (3) which has been
the first change on the course since the reunion happened. So the course in its current state
is represented as C(A3,1) T5.
22