Content uploaded by Dietrich Albert
Author content
All content in this area was uploaded by Dietrich Albert
Content may be subject to copyright.
H. Leung et al. (Eds.): ICWL 2007, LNCS 4823, pp. 78 – 89, 2008.
© Springer-Verlag Berlin Heidelberg 2008
The ELEKTRA Ontology Model:
A Learner-Centered Approach to Resource Description
Michael D. Kickmeier-Rust and Dietrich Albert
Cognitive Science Section, Department of Psychology, University of Graz
Universitätsplatz 2 / III, 8010 Graz, Austria
{michael.kickmeier, dietrich.albert}@uni-graz.at
Abstract. There is little doubt that intelligent and adaptive educational tech-
nologies are capable of providing personalized learning experiences and im-
proving learning success. Current challenges for research and development in
this field concern, for example, the design of comprehensive data models for
adaptive systems as well as the interoperability of systems and the re-
usability of learning material across different systems. In the present work we
introduce an ontology model, basically developed in the context of immersive
digital games, which attempts to provide a solution to existing problems in
resource description. On the one hand, comprehensive data models for adap-
tive systems are supported by separating static information from adaptive
systems as far as possible. On the other hand, the ontology model offers a po-
tential solution to precise and, above all, learner-centered resource description
by separating latent competencies from observable performance (in learning
objects or test items).
Keywords: Adaptive Tutoring, Game-based Learning, Resource Description,
Ontology Model.
1 Introduction
In the past two decades learning technologies dramatically changed. Web-based solu-
tions for educational purposes are widely accepted and commercially successful.
Intelligent and adaptive educational technologies are a logical evolution, acknowledg-
ing the need for tailoring individual learning experiences. However, such intelligent
and adaptive systems did not open up the market yet. Reasons are seen in the diffi-
culty of designing comprehensive data models, interoperability, and re-usability of
learning media as well as in a lack of a focus on the learner. At the same time, excit-
ing new technologies for (web-based) learning are already dawning, for example
game-based learning. The emergence of these new technologies is attended by further
challenges and it is increasing the need for precise data models and resource descrip-
tion frameworks. In the present work we introduce an ontology model, developed in
the context of game-based learning, that attempts to provide a solution to existing
problems of adaptive systems as well as a potential solution to precise and, above all,
learner-centered resource description.
Kickmeier-Rust, M. D., & Albert, D. (2008). The ELEKTRA ontology model: A learner-
centered approach to resource description. Advances in Web Based Learning – ICWL
2007 (pp. 78-89). Lecture Notes in Computer Science, 4823/2008. Berlin: Springer.
The ELEKTRA Ontology Model 79
1.1 Intelligent and Adaptive Tutoring
The idea of using “intelligent” machines for educational purposes has a long tradition.
It can be traced back to 1926 when Pressey [1] tried to build a machine that presented
multiple choice questions and immediate feedback on the answers. Psychologists and
educationists have since reported that carefully designed individualized tutoring pro-
duces the best learning for most people (e.g., [2]). First so-called intelligent tutoring
systems (ITS) for web-based application were reported in the mid-nineties [3] and
also research on adaptive hypermedia turned towards educational objectives, develop-
ing adaptive tutoring systems (ATS). Both use similar approaches and techniques to
realize individualized tutoring.
Adaptive presentation refers to providing individual learners with personalized in-
formation, for example by conditional inclusion of information, re-ordering of informa-
tion, or providing different media types [4]. Adaptive navigation support refers to guid-
ing an individual learner through the learning material in the most suitable and
successful way, for example by direct guidance, link sorting, or link hiding. Problem
solving support is basically a concept of ITS that attempts to provide a learner not only
with the final solution of a problem, for example when the learner is stuck, but to ana-
lyze how a solution was obtained and which knowledge might be missing, or which
misconceptions might have been the cause for an error. An alternative terminology
comes from [5], basically established in the context of educational games, macro and
micro adaptivity. Essentially, macro adaptivity refers to traditional approaches of adap-
tive presentation and navigation on the level of LO or learning situations (LeS) whilst
micro adaptivity refers to adaptive presentation and problem solving support within
a LO/LeS. Although there is a strong need for personalized tutoring, such techno-
logies are implemented sparsely in commercial E-Learning platforms. Reasons are for
example:
• Adaptive features might still be a less important and visible factors in the
market [6]
• ITS/ATS are often technology-driven and lack a plausible psychological
and pedagogical background [6]
• Although there are significant efforts spent on providing suitable resource
description [7], these are not commonly accepted and probably not suffi-
ciently powerful to describe learning resources, especially in the context
of ATS
• A number of authors [8] argue that it still is difficult to re-use and ex-
change learning material across different applications because of the often
strong concatenation of LO, adaptive logic, and psycho-pedagogical
background
• Most often the focus of resource description is -quite naturally- on LO,
having the disadvantage of ambiguity of different learning methods and
different learning objects covered by a single LO
The development of ITS/ATS is still facing major challenges and existing resource
description approaches and standards are not commonly accepted yet and may be not
complete enough. At the same time, the “very next big thing” (following the title of
80 M.D. Kickmeier-Rust and D. Albert
Paul De Bra’s article “The next big thing: Adaptive web-based systems” [9]) that is
already dawning is using immersive digital games for educational purposes.
1.2 The Very Next Big Thing: Immersive Digital Educational Games
The majority of E-Learning systems and multimedia LO are based on traditional 2D
user interfaces; provocatively speaking, they have all more or less the same unexcit-
ing look and feel. This perspective is compounded by the proliferation of appealing
computer games. The idea of using such games for educational purposes was already
born with the appearance of the first computer games. Ever since, scientists and de-
velopers have published numerous articles and books on the advantages of digital
game-based learning as a promising approach to improve and facilitate learning, es-
pecially when fun, motivation, and immersion could be maintained [10]. In addition
to single player games for educational purposes, multiplayer online games increas-
ingly get in the focus of educational research [11]. Such games are interesting from a
psycho-pedagogical perspective because they incorporate possibilities for collabora-
tive peer-to-peer learning and social interactions.
However, educational games, and especially educational multiplayer games, bear
further challenges to adaptive technologies, underlying data models, and resource
description models:
• The complexity and scale of LO/LeS is substantially higher than in tradi-
tional LO. Moreover, LO/LeS are strongly integrated in a specific game’s
narrative and visual style
• The costs of developing LO/LeS for immersive, state-of-the-art educa-
tional games are extremely high and successful approaches to re-usability
are even more important than in traditional E-Learning
• Adaptive technologies are facing new challenges in order to provide suit-
able adaptive interventions
• In multiplayer games learning may be independent from pre-described
LO/LeS, for example through collaboration, peer-tutoring, and social in-
teractions.
2 The ELEKTRA Project
The ELEKTRA project (www.elektra-project.org) has the ambitious goal to utilize
the advantages of computer games and their design fundamentals for educational
purposes and to address disadvantages of game-based learning as far as possible.
Within the project a methodology for successful design of educational games shall be
established and a game demonstrator is developed based on a state-of-the-art 3D ad-
venture game teaching optics according to national curricula. ELEKTRA will also
address important research questions concerning data model design as basis for adap-
tivity and resource description enabling interoperability of systems and re-using
LO/LeS as well as the data model itself.
In view of the mentioned challenges to adaptive technologies, data models, and re-
lated resource description frameworks and also in view of the emerging challenges by
The ELEKTRA Ontology Model 81
(multiplayer) educational games, we propose a conceptual change towards a separa-
tion of competence (i.e., a set of skills) and performance.
2.1 Competence versus Performance
An early distinction between latent competence and observable performance was
introduced by Chomsky [12] in the framework of linguistic theory. Today, this dis-
tinction has a much wider application, especially in knowledge and learning psychol-
ogy. Still, in practice the concepts of latent competence and related observable
performance often lack a thorough differentiation; operationalizations are often one-
to-one mappings of underlying competencies and performance and often the same
labels are used for both concepts. From a cognitive point of view, this approach is
fraught with difficulty; for example, it does not acknowledge that performance (e.g.,
mastering a task) can be the result of various competencies. Thus, it is not only neces-
sary to break down certain types of competencies to a certain level of granularity but
also to separate competence from performance. Such separation enables establishing a
sound basis to address existing challenges: (a) it offers a learner-centered and cogni-
tively sound approach, (b) it enables resource description (and probably standards)
without the focus on specific LO; it is not about which content is included in a LO,
but what exactly a learner can gain from a LO, and (c) it enables a separation of LO,
adaptive mechanism, and psycho-pedagogical principles and, therefore, serves the
design of adequate data models underlying adaptive systems.
A method to realize such separation is ontologies. On such basis, a clear, precise,
and probably standardized definitions of competencies in a given domain can be real-
ized which, in turn, can be used as a data model for an adaptive system. The cognitive
model underlying the data model of ELEKTRA is based on Competence-based
Knowledge Space Theory (CbKST).
To address the challenges for research and development and to incorporate a sepa-
ration of latent competence and observable performance, ELEKTRA utilizes the
framework of CbKST to provide the game with a methodology for suitable adaptive
interventions. It provides an internal cognition-based logic that is quite similar to the
logic of ontologies: well-defined entities (the competencies or skills) are in a well-
defined relationship (a so-called prerequisite relation).
CbKST is an extension of the originally behavioral Knowledge Space Theory [13]
[14] where a knowledge domain Q is characterized by a set of problems. The knowl-
edge state of an individual is identified on the subset of problems this person is capa-
ble of solving. Due to mutual dependencies between the problems captured by
prerequisite relations, not all potential knowledge states will occur. The collection of
all possible states is called a knowledge structure K. To account for the fact that a
problem might have several prerequisites (i.e., and/or-type relations) the notion of a
prerequisite function was introduced. The basic idea of CbKST is to assume a set E of
abstract skills underlying the problems and LO of the domain. The relationships be-
tween the skills and problems (or LO) is established by a skill function. Such function
assigns a collection of subsets of skills (i.e., competence states) to each problem,
which are relevant for solving it and it assigns the skills to each LO taught. By associ-
ating skills to the problems of a domain, a knowledge structure on the set of problems
82 M.D. Kickmeier-Rust and D. Albert
Fig. 1. The left panel illustrates a prerequisite function (the bended line below skill X indicates
a logical or). The right panel shows the corresponding competence structure. The bolded line
indicates one of several meaningful learning paths.
is induced. The skills, which are not directly observable, can be uncovered on the
basis of a person’s observable performance. A further extension is to assume prereq-
uisite relationships between the skills, inducing a competence structure C on the set of
skills [15]. To illustrate this approach, assume that a knowledge domain is represented
by Q={a, b, c, d}. Consider the set E={V, W, X, Y, Z} of skills that are relevant for
solving them. A prerequisite function that might exist among these skills is demon-
strated in Fig. 1a. For example, this function reads that if a student has skill X we can
assume that this student also possesses either skill V or W, or both; the corresponding
competence structure is shown in Fig. 1b. It includes only 13 possible competence
states from a total of 25 = 32 states.
The outlined approach entails several advantages. Given the performance, that is, the
subset of problems a student could master, the latent skills underlying that problem solv-
ing performance can be identified. Due to the utilization of representation and interpreta-
tion functions no one-to-one mapping of performance to skills is required and meaningful
learning paths can be identified.
3 The ELEKTRA Ontology Model
3.1 Ontology Web Language (OWL)
Originally, the term ontology was established in philosophy where it describes a dis-
cipline concerned with existence. The term was introduced to computer science by
Gruber [16] where it describes “a formal, explicit specification of a shared conceptu-
alization”. An ontology provides a structured and semantically rich approach to model
a certain domain; ontologies count classes, instances, inheritances, and relationships
between classes as their major components. The entities of an ontology can be de-
scribed with attributes, each having a name, a certain data type, and one or more val-
ues. In the context of E-Learning, ontologies serve as a means of achieving semantic
precision between a domain of learning material and the learner’s prior knowledge
and learning goals. Ontologies bridge the semantic gap between humans and ma-
chines and, consequently, they facilitate the establishment of the semantic web and
The ELEKTRA Ontology Model 83
build the basis for the exchange and re-use of contents that reaches across people and
applications. From a technical perspective, an ontology is a text-based reference of
information, represented by an ontology representation language. Most of them are
built upon XML and RDF. There is a variety of such representation languages (see
[17] for a review). The primary benefit of using ontologies is their ability to reason
over defined relationships and therefore to relate instances to their abstract types.
Reasoning is used to derive new relations between individuals from an existing ontol-
ogy by using and applying logical rules. Reasoning might refer to class memberships,
to the equivalence of classes, to consistency, or to classifications. As an example,
ontologies allow determining a complete list of skills required by a specific compe-
tence state or by a specific LO.
In 2004, W3C has officially released OWL (Web Ontology Language) as recom-
mendation for representing ontologies. OWL is developed starting from description
logic (DL) and DAML+OIL. The popularity of OWL, which is still increasing, might
lead to its establishment as the standard ontology representation language on the se-
mantic web. Basically, OWL is a set of XML entities and attributes with well-defined
meaning that are used to identify concepts and relations between. OWL consists of
three species, OWL Full, OWL DL, and OWL Lite [23]. OWL provides a set of con-
structors (e.g., oneOf, intersectionOf, hasValue) that allow deriving classes from other
classes and a set of axioms (e.g., subClassOf, disjointWith, sameAs, TransitiveProp-
erty) that allow asserting subsumption or equivalence in terms of classes, individuals,
or properties, the disjointness of classes, or properties of properties.
3.2 Architecture
The ELEKTRA ontology model is supposed to address the requirements of providing
adaptive interventions in the game in order to balance challenge and ability and there-
fore not only providing successful learning paths within the game’s narrative but also
to retain motivation and even flow experience. Thus, the ontology model incorporates
the concepts related to adaptive interventions on a macro as well as on a micro-level,
that is, it models problem solution spaces for problem solving tasks within the game
environment. Although the re-usability of learning objects is more difficult in digital
games for educational purposes because the learning objects are an integral part of the
entire game which is not interchangeable between different games in most cases, the
presented ontology model is supposed to serve the growing demands on standardiza-
tion and semantically rich resource description in the context of educational technolo-
gies also. As mentioned before, some authors argue that currently a lack of commonly
accepted resource description standards for learning objects exists. One reason might
be a focus on learning objects in current approaches. In the presented ontology model
we introduce a focus on the learner and, therefore, on latent skills. This approach
might be a more comprehensive and easier to standardize method for describing and
defining LO and learning objectives. While learning objects are most often strongly
interlaced with instructional methods or events, the focus on underlying skills offers
a cornerstone, which is not only directly related to human abilities and learning
84 M.D. Kickmeier-Rust and D. Albert
Fig. 2. The ELEKTRA ontology model architecture
objectives but which also allows a very precise description of LO. From a technical
perspective the ELEKTRA ontology model builds upon OWL DL; due to its popular-
ity it is the quasi-standard for ontology representation languages. All classes have the
RDF-attributes label and description. The ontology model is illustrated in Fig. 2. Most
relations between the classes’ instances are non-functional (marked with an asterisk in
Fig. 2), meaning that one or more instances of one class can be associated with the
instances of a related class.
The ELEKTRA Ontology Model 85
Learner. The ELEKTRA ontology model puts the human learner and, therefore,
skills in the foreground. Consequently Learner is the center class of the ontology.
Using attributes such as age, school level, sex, culture, country, or learning styles
allows establishing distinct groups of learners. In turn, such well-defined groups of
learners provide the adaptation engine of a learning environment, the game by the
example of ELEKTRA, with a comprehensive learner model. The Learner class is
(indirectly) associated with classes related to latent skills, LO (LeS in the terminology
of ELEKTRA), problem solution states, and curriculum.
Skills. The most important component describing the learner is skills. The Skill class
is defined by a factual concept (e.g., convex lenses) and related action verbs (e.g.,
recall or apply). This type of skill definition was introduced by [18] and is associated
with Bloom’s revised taxonomy of learning objectives [19]. There are six cumulative
levels of cognitive processing, which can be thought of as degrees of difficulties
which establish a hierarchical order; a more simple level of knowledge or ability must
be given in order to reach a deeper one. The levels are (a) knowledge (the recall of
factual information), (b) comprehension (understanding of the meaning, translation,
interpolation, and interpretation of instructions and problems), (c) application (using
of a concept in a new situation or unprompted use of an abstraction), (d) analysis
(separating material or concepts into sub- components so that its organizational
structure may be understood), (e) synthesis (building a structure or pattern from
diverse elements, joining parts to form a whole, with emphasis on creating a new
meaning or structure), and (f) evaluation (making of judgments about the value of
ideas or materials). Action verbs are assigned to each category, describing recall
methods or knowledge more detailed. For the category “knowledge” action verbs are,
for example, “define”, “describe”, or “label”; for the category “analysis” “compare”,
“quantify”, or “measure”. The Skill class has a relational attribute has_prerequisite.
This relation identifies skills that are prerequisites for a given skill as claimed by
CbKST (see section 2.2) and, therefore, establishes a prerequisite relation between the
skills. Fig. 3 shows the prerequisite relation for the physics course realized in the
ELEKTRA game demonstrator. A related class is the SkillSet class. This class
specifies a set of skills, which is in turn a prerequisite for other skills. This
supplement is based on rather technical constraints when including and/or-type
relations (i.e., prerequisite functions).
Learning Objects/Learning Situations. A further class defines LO or, in the
context of game-based learning, LeS. The LearningSituation class has the relational
attribute skills_taught, which refers to the skills that can be learned with certain LO or
within a certain LeS. To acknowledge the fact that certain skills may be required to
successfully apply such LO/LeS this class has the relational attributes skills_required
and skillsets_required. The present ontology model includes both because the
definition of skill sets is necessary when and/or-type prerequisite functions are
required.
Although learning and assessment is often overlapping, in the present model we
distinguish between LO/LeS and assessment objects/situations. Such objects are
86 M.D. Kickmeier-Rust and D. Albert
Fig. 3. Upward drawing for the prerequisite relation between skills realized in ELEKTRA
typical test items or learning situations primary aiming at assessing knowledge with
which the current knowledge state is assessed. The separation facilitates the adaptive
presentation of learning material and assessment of learning progress. The Assess-
mentSituation class has the attributes skills_required and skillsets_required, specify-
ing the skills that are necessary to successfully master such assessment. To account
for pedagogical implications and strategies, two related classes concern the type of
learning events and the depth of knowledge. Learning events (prototypically) refer to
the Eight Learning Events Model [20], which is a pedagogical approach emphasizing
that learning events are based on eight basic components. The eight learning events
are (a) imitation / modeling, (b) reception / transmission, (c) exercising / guidance, (d)
exploration / documentation, (e) experimentation / reactivity, (f) creation / confronta-
tion, (g) self-reflection / co-reflection, and (h) debate / animation. The depth of
knowledge refers to Bloom’s taxonomy of learning objectives. To include these
classes, the LearningSituation class has the attributes covers_event and covers_depth
and the AssessmentSituation class has the attribute covers_depth.
Curricula and Units. An aim of ELEKTRA is to design a methodology for game-
based learning that is close to school curricula. The game demonstrator, for example,
will be relying on the curricula of France, Belgium, and Germany. To integrate such
information, we used a Curriculum class. This class has the attributes subject, level,
country and release (version or date) to identify the curriculum. To link the LO/LeS
to the curricula, this class has also the relational attribute includes_learningsituations.
A similar class is LearningUnit, which specifies larger learning units within the game
or curriculum.
Problem Solution States. To provide a basis for micro adaptivity (as described in
section 1.1), which attempts to analyze the states of a problem solution process in order
to provide subtle adaptive interventions (e.g., giving hints), we included the Object class.
This class is strongly related to the game-approach and identifies the manipulable objects
that exist in different learning and assessment situations (e.g., books, tools, lenses,
microscopes, etc.). These objects are linked to learning and assessment situations by the
exists_in attribute. In order to allow the adaptive system to determine to progress in the
The ELEKTRA Ontology Model 87
problem solution process (and also possible misconceptions), this class is linked to
the Position_Category class. The attributes poscat_value, poscat_skills_missing, and
poscat_skills_required enable the assignment of some value of correctness, determined
by a specific utility function, to each position category of each manipulable object and it
allows determining available and missing skills. On the basis of the correctness value the
probability distribution of the related skills states can be updated and interventions can be
made on a skill-basis rather than on behavioral basis.
Domain Ontologies. As included in Fig. 2, in addition to the ontology structure that
serves as a data model for the adaptive system, either on a macro or on a micro-level,
also a domain ontology can be linked to the data model. Domain ontologies generally
include propositions (two factual concepts which are connected by any type of
relation) that describe a certain domain (e.g., the domain of optics). Such domain
ontologies can be used to enable a manual or semi-automatic derivation of skills and
the prerequisite relations between them [21].
4 Conclusion and Future Work
In the present work we introduced an ontology-based data model that essentially was
developed in the context of game-based learning. This model is anchored in CbKST
and supposed to enable suitable adaptive interventions on a macro and on a micro-
level. In the framework of the ELEKTRA project, such adaptive intervention not only
involves the learner’s knowledge and learning progress, it involves pedagogical im-
plications and strategies as well as motivational characteristics and immersion. Al-
though developed in view of a specific application, the data model might serve as a
role model for other adaptive systems and approaches to adaptive interventions.
The primary aim of the introduced ontology model is providing adaptive systems
with a basis for suitable adaptive interventions and to separate static information from
the adaptive system as far as possible. The architecture enables the realization of the
initially described techniques of adaptivity (adaptive presentation, adaptive naviga-
tion, and adaptive problem solving support), on a macro as well as on a micro-level.
In a context of probabilistic skill assessment and, particularly, in the context of evalu-
ating the “correctness” of problem solutions states, which includes some characteris-
tics of fuzzy logic, existing ontology representation languages have limitations. For
example in OWL any sentence (e.g., reasoning results) must be either true or false.
Consequently, ontologies cannot quantify the degree of the overlap or inclusion of
concepts [22]. Future endeavors will incorporate existing approaches to probabilistic
extensions to ontology representation languages (e.g., the Bayesian network approach
of [22]).
In addition, we proposed a conceptual change towards a learner-centered focus on
latent skills in resource description methods. Both data model and competence-
performance separation may offer a promising approach to address the problems and
challenges of adaptive educational technologies. First, the introduced approach to
separate competence and performance and to include this separation in the data model
offers a cognitive basis opening the doors for new development in knowledge con-
struction and assessment. Furthermore, the data model acknowledges the need for
88 M.D. Kickmeier-Rust and D. Albert
pedagogical implications and strategies and for the integration of individual states
and traits (e.g., motivational components). On the basis of this triad, LO/LeS and
learning objectives can be defined very precisely. In view of resource description
methods and their standards, the proposed separation of a learner’s possessed and
desired skills from LO and learning objectives establish similar to propositions of
domain ontologies smallest - or at least sufficiently small - entities for describing the
learner, LO/LeS, and learning objectives. At the same time, problems in tracking
learning progress emerging from multi-learner environments (such as multi-player
games), for example learning by social interactions and independent from specific
LO, can be reduced by swerving from a focus on LO/LeS. Future endeavors will
extent the present ontology model towards an increasing compatibility with existing
attempts to provide metadata standards and to include ontologies for the definition of
skills or competencies (e.g., the IMS information model specification for reusable
definition of competency or educational objective). Moreover, the presented model
offers a suitable supplement to attempts of combining learning design standards with
ontology models, for example by providing a generalized taxonomy of pedagogical
strategies and learning objectives or by acknowledge information for adaptive prob-
lem solving support.
Acknowledgements. The research and development introduced in this work is funded
by the European Commission under the sixth framework programme in the IST
research priority, contract number 027986.
References
1. Pressey, S.L.: A simple apparatus which gives tests and scores - and teaches. School and
Society 23, 373–376 (1926)
2. Bloom, B.S.: The 2 sigma problem: The search for methods of group instruction as affec-
tive as one-to-one tutoring. Educational Researcher 13(6), 4–16 (1984)
3. Brusilovsky, P.: Intelligent tutoring systems for World-Wide-Web. In: 3rd International
WWW Conference (1995)
4. Brusilovsky, P.: Adaptive and intelligent technologies for web-based education. In: Roll-
inger, C., Peylo, C. (eds.) Special Issue on Intelligent Systems and Teleteaching, Kün-
stliche Intelligenz, vol. 4, pp. 19–25 (1999)
5. Kickmeier-Rust, M.D., Peirce, N., Conlan, O., Schwarz, D., Verpoorten, D., Albert, D.:
Immersive digital games: The interfaces for next-generation E-Learning? In: Aykin, N.
(ed.) HCII 2007. LNCS, vol. 4560, Springer, Heidelberg (2007)
6. Branko Neto, W.C., Gauthier, F.A.: Sharing and reusing information on web-based learn-
ing. In: Workshop on Applications of Semantic Web Technologies for E-Learning (2006)
7. Mizoguchi, R., Bourdeau, J.: Using ontological engineering to overcome common AI-ED
problems. International Journal of AI in Education 11, 107–121 (2000)
8. Sampson, D., Karagiannidis, C., Cardinali, F.: An architecture for web-based E-Learning
promoting re-usable adaptive educational e-content. Educational Technology & Soci-
ety 5(4), 27–37 (2002)
9. De Bra, P., Aroyo, L., Chepegin, V.: The next big thing: Adaptive web-based systems.
Journal of Digital Information 5 (2004)
The ELEKTRA Ontology Model 89
10. Kickmeier-Rust, M.D., Schwarz, D., Albert, D., Verpoorten, D., Castaigne, J.-L., Bopp,
M.: The ELEKTRA project: towards a new learning experience. In: Pohl, M., et al. (eds.)
M3 – Interdisciplinary aspects on digital media & education, pp. 19–48 Österreichische
Computer Gesellschaft, Vienna (2006)
11. Steinkuehler, C.A.: Massively multiplayer online videogaming as participation in a dis-
course. Mind, Culture, & Activity 13(1), 38–52 (2006)
12. Chomsky, N.: Aspects of the Theory of Syntax. MIT Press, Cambridge (1965)
13. Doignon, J.-P., Falmagne, J.-C.: Knowledge spaces. Springer, Heidelberg (1999)
14. Albert, D., Lukas, J.: Knowledge spaces: Theories, empirical research, and applications.
Lawrence Erlbaum, Mahwah (1999)
15. Korossy, K.: Modelling knowledge as competence and performance. In: Albert, D., Lukas,
J. (eds.) Knowledge Spaces: Theories, Empirical Research Applications, pp. 103–132.
Lawrence Erlbaum, Mahwah (1999)
16. Gruber, T.A.: A translation approach to portable ontology specifications. Knowledge Ac-
quisition 5, 199–220 (1993)
17. Gómez-Pérez, A., Corcho, O.: Ontology languages for the semantic web. IEEE Intelligent
Systems 17, 54–60 (2002)
18. Heller, J., Steiner, C.M., Hockemeyer, C., Albert, D.: Competence-based knowledge struc-
tures for personalized learning. International Journal of E-Learning 5, 75–88 (2006)
19. Anderson, L.W., Krathwohl, D.R., et al.: A taxonomy for learning, teaching and assessing:
A revision of Bloom’s Taxonomy of educational objectives: Complete edition, Longman
(2001)
20. Leclerq, D., Poumay, M.: The 8 Learning events model and its principles, http://www.
labset.net/media/prod/8LEM.pdf
21. Mayer, B., Steiner, C.M., Heller, J., Albert, D.: Activity- and taxonomy-based knowledge
representation framework. International Journal of Knowledge and Learning (in press)
22. Ding, Z., Peng, Y.: A probabilistic extension to ontology language OWL. In: 37th Interna-
tional Conference on System Sciences (2004)
