ArticlePDF Available

Abstract

Competence structures of the content and competence modelling of the learners provide well-suitable means for finding appropriate peer tutors in CSCL based on asynchronous messaging and annotations. Various criteria for the appropriateness of potential peer tutors can be discussed. An internet-based system used as a large scale Web experiment can then also deliver data for evaluating such criteria.
Applying Competence Structures for Peer Tutor Recommendations in CSCL
Environments
Jürgen Heller, Cord Hockemeyer, Dietrich Albert
University of Graz, Austria
Abstract
Competence structures of the content and compe-
tence modelling of the learners provide well-suitable
means for finding appropriate peer tutors in CSCL
based on asynchronous messaging and annotations.
Various criteria for the appropriateness of potential
peer tutors can be discussed. An internet-based sys-
tem used as a large scale Web experiment can then
also deliver data for evaluating such criteria.
1: Introduction
An important issue in CSCL is the selection of peer
tutors or other co–learners for a student. In this paper, we
will focus on peer tutoring based on public annotations to
documents in an e–Learning course delivered through the
Web. This approach connects Hoppe’s [10] formalisation
of peer selection criteria and Inaba and Okamoto’s [11,
12] concept of utterance classification with the theory of
knowledge spaces which facilitates personalised e–
Learning through the application of prerequisite struc-
tures.
2: Theoretical background
2.1: Peer selection
Hoppe [10] has formally described the selection of
collaboration groups based on the student modelling
within an e–Learning system and on the system’s knowl-
edge about the problem to be solved within the group. In
case of peer tutoring, the tutor should be able to solve the
tasks whereas the tutored person has some difficulties. In
case of collaborative problem solving, all necessary
knowledge should be available in the group while no
member has all needed knowledge on their own.
2.2: Utterance classification
Inaba and Okamoto [11, 12] have presented an e–
Learning system which allows the learners to add their
utterances (questions, answers, comments, etc.) as classi-
fied, public annotations to an existing course. A learner
publishing such an utterance has to specify the type of it.
In case of comments, this includes also the specification
whether the comment is an agreement or a contradiction.
2.3: Knowledge space theory
The theory of knowledge spaces was developed by
Doignon and Falmagne [5,6] originally aiming at the
adaptive assessment of knowledge. In the meantime, how-
ever, its major application lies in the field of e–Learning.
If a domain of knowledge is described through a set
of test items, there will often exist prerequisite relation-
ships between these items, i.e. if a student is capable of
solving a certain item a, we can surmise that this student
is also capable of solving some other item b. Such prereq-
uisite structures can be used for inferences for adaptive
testing. If a prerequisite structure contains also lessons, we
can use the lessons and items together with the structure
for personalised e–Learning [2]. This approach has been
applied in the development of the RATH (Relational
Adaptive Tutoring Hypertext) system [9].
In an extension of knowledge space theory, the group
around Albert and Lukas [1, 3] introduced a clear distinc-
tion between concrete learning objects and abstract, latent
competencies (or skills). Such a distinction supports the
development of adaptive e–Learning systems with highly
reusable content including the structure information
through metadata usage [4,8].
3: The C²RATH concept
3.1: Collaborative RATH (CRATH)
Based on the experiences with RATH and on the re-
sults of Hoppe and of Inaba and Okamoto, the concept of
a collaborative RATH system was developed. In RATH,
the knowledge of a learner is modelled as the set of learn-
ing documents the learner has read and the set of items
s/he has solved [9].
In the collaborative RATH concept study [7], a learner
would be able to add public annotations to a document.
Such an annotation would be classified as described in
Section 2.2. In case of a question, the system would select
some other learners with an appropriate knowledge state
and ask for their willingness to answer the question. On
the other side, whenever some learner gives an annotation
to an existing annotation, the author of that original anno-
Proceedings of the IEEE International Conference on Advanced Learning Technologies (ICALT’04)
0-7695-2181-9/04 $20.00 © 2004 IEEE
tation would be notified.
One factor in the selection of appropriate co–learners
would also be the online–status, i.e. a potential peer who
is currently online would be preferably selected in order to
ensure a rapid answer to a question.
The appropriateness of a peer may well have criteria
beyond those of Hoppe. One might, e.g., prefer peers who
share much knowledge with the asking learner. This
would allow for analogies drawn by the peer to be under-
stood by the asking student.
3.2: Collaborative, Competence–based RATH
(C²RATH)
One disadvantage of the RATH system is its docu-
ment-focused approach. Whenever the documents in a
course change (including the addition of new and the
elimination of existing documents), prerequisite informa-
tion for other, unchanged documents may also change.
This makes it quite difficult to maintain a RATH course
[4].
The APeLS system [4, 8] provides a solution for this
through the separation of learning objects and abstract
concepts or competencies. For a learning object in APeLS,
the prerequisite objects are specified indirectly through
prerequisite competencies.
Introducing the competence approach into the
CRATH idea would provide the same advantages as the
development from RATH to APeLS. The student model
includes the competencies acquired by the respective
learner. However, for selecting appropriate peers, the set
of documents read by both learners would still be impor-
tant: the peers selected do not necessarily have to have
read the very document under discussion. Actually, it will
often be helpful if a peer has (also) another access to the
competencies taught in that document.
4: Discussion
We have proposed an approach for peer selection in a
competency–based learning environment. This approach
has, of course, to be implemented and tested for its feasi-
bility in practice.
Once such a system exists and its feasibility has been
proven, it may furthermore be used to investigate different
criteria for peer selection. So far, there has been not much
research on which criteria are best for this selection, and a
system open for varying them can be an important re-
search tool into this direction.
Acknowledgements
Part of this work was supported by the European
Commission through a Marie Curie Fellowship to the
second author (Grant no. ERBFMBICT983377).
References
[1] Albert, D. (Ed.). (1994). Knowledge structures. New York:
Springer Verlag.
[2] Albert, D., & Hockemeyer, C. (2002). Applying Demand
Analysis of a Set of Test Problems for Developing an Adaptive
Course. Proceedings of the International Conference on Com-
puters in Education ICCE 2002 (pp. 69–70). Los Alamitos:
IEEE Computer Society.
[3] Albert, D. & Lukas, J. (Eds.). (1999). Knowledge spaces:
Theories, empirical research, applications. Mahwah, NJ: La-
wrence Erlbaum Associates.
[4] Conlan, O., Hockemeyer, C., Wade, V., & Albert, D. (2002).
Metadata Driven Approaches to Facilitate Adaptivity in Person-
alized eLearning systems. The Journal of Information and Sys-
tems in Education, 1, 38–44.
[5] Doignon, J.-P. & Falmagne, J.-C. (1985). Spaces for the
assessment of knowledge. International Journal of Man-
Machine Studies, 23, 175–196.
[6] Doignon, J.-P. & Falmagne, J.-C. (1999). Knowledge
spaces. Berlin: Springer_Verlag.
[7] Hockemeyer, C. (2000). CRATH: A Collaborative Adaptive
Tutoring Hypertext System (Unpublished Technical Report).
Institut für Psychologie Karl–Franzens–Universität Graz,
Austria.
[8] Hockemeyer, C., Conlan, O., & and Albert, V. W. D. (2003).
Applying Competence Prerequisite Structures for eLearning and
Skill Management. Journal of Universal Computer Science, 9,
1428–1436.
[9] Cord Hockemeyer, Theo Held, and Dietrich Albert. RATH
— a relational adaptive tutoring hypertext WWW–environment
based on knowledge space theory. In Christer Alvegård, editor,
CALISCE`98: Proceedings of the Fourth International Confer-
ence on Computer Aided Learning in Science and Engineering,
pp. 417-423, Göteborg, Sweden, June 1998. Chalmers
University of Technology.
[10] Hoppe, H. U. (1995). The use of multiple student modeling
to parameterize group learning. In J. Greer (Ed.), Artificial Intel-
ligence in Education, 1995 (pp. 234–241). Charlottesville, VA:
Association for the Advancement of Computing in Education
(AACE).
[11] Inaba, A. & Okamoto, T. (1995). The network discussion
supporting system embedded computer coordinator at the dis-
tributed places. Educational Technology Research, 18, 17–24.
[12] Inaba, A. & Okamoto, T. (1997). The intelligent discussion
coordinating system for effective collaborative learning. In T.
Okamoto & P. Dillenbourg (Eds.), Collaborative Learn-
ing/Working Support System with Networking (pp. 26–33). Kobe,
Japan. Workshop at the 8th World Conference on Artificial Intel-
ligence in Education AI–ED 97.
Proceedings of the IEEE International Conference on Advanced Learning Technologies (ICALT’04)
0-7695-2181-9/04 $20.00 © 2004 IEEE
... It has aroused a great deal of interest in the field of education for a variety of reasons. First, it is a powerful instructional strategy for inclusive education (Ainscow, 1991), second, it fosters the skills and attitudes that are fundamental to building a democratic society to be constructed (Slavin, 1995) and finally it constitutes one of the pillars of networked learning (Heller, Hockemeyer, & Albert, 2004). It is also an excellent resource for promoting the mastering of the interpersonal competencies that are so crucial in the society of knowledge. ...
Article
Full-text available
The research analyzes the interaction of 24 students (12 pairs) of secondary students when using peer tutoring techniques to learn Catalan. Students worked together in a program to produce an authentic writing experience. Significant increases were observed in pre-and posttest Catalan attainment scores of students. An analysis of the interactivity, or joint activity highlighted the emergence of 2 patterns: the tutor's active pattern and the tutee's reactive pattern. Different styles of cooperative interactions were also identified, tutors tended to utilize cooperative messages, and tutees tended to initiate collaborative messages. A sequential analysis of the exchanges in the peer tutoring session identified 3 types of sequences: collaborative, tutorial and the initiation-response-feedback.
Article
An intelligent tutoring system for learning basic statistics, called Stat-Knowlab, is presented and analyzed. The algorithms implemented in the system are based on the competence-based knowledge space theory, a mathematical theory developed for the formative assessment of knowledge and learning. The system’s architecture consists of the two assessment and learning modules that interact with each other in a continuous exchange of information about the current knowledge state of a student. This allows the system to personalize the student’s learning, providing only with the learning objects that she is ready to learn. During the browsing of the system, several types of navigation data are recorded. In this work, we analyzed data from two studies that were aimed at examining the learning processes induced by the navigation of the system. The results of both studies highlighted that the system is useful for monitoring the student learning processes during a university course of basic statistics.
Conference Paper
Full-text available
Assessing a learner's mastery of a set of skills is a fundamental issue in intelligent learning environments. We compare the predictive performance of two approaches for training a learner model with domain data. One is based on the principle of building the model solely from observable data items, such as exercises or test items. Skills modelling is not part of the training phase, but instead dealt with at later stage. The other approach incorporates a single latent skill in the model. We compare the capacity of both approaches to accurately predict item outcome (binary success or failure) from a subset of item outcomes. Three types of item-to-item models based on standard Bayesian modeling algorithms are tested: (1) Naive Bayes, (2) Tree-Augmented Naive Bayes (TAN), and (3) a K2 Bayesian Classi er. Their performance is compared to the widely used IRT-2PL approach which incorporates a single latent skill. The results show that the item-to-item approaches perform as well, or better than the IRT-2PL approach over 4 widely di erent data sets, but the di erences vary considerably among the data sets. We discuss the implications of these results and the issues relating to the practical use of item-to-item models.
Article
Full-text available
Personalized eLearning Systems tailor the learning experience to characteristics of individual learners. These tailored course offerings are often comprised of discrete electronic learning resources, such as text snippets, interactive animations, diagrams, and videos. An extension of standard metadata schemas developed for facilitating the discovery and reuse of such adaptive learning resources can also be utilized by the eLearning systems for realizing the adaptivity. An important feature of such reuse supporting adaptive systems is the clear distinction of separate models and components within the teaching process.
Article
Full-text available
This paper investigates how technological distance between firms affects their network of R&D alliances. Our theoretic model assumes that the benefit of an alliance between two firms is given by their technological distance. This benefit-distance relationship determines the ego-network of each firm as well as the overall network structure. Empirical relevance is confirmed for the bio-pharmaceutical industry. Although we find that the network structure is largely explained by firm size, technological distance determines the positioning of firms in the network.
Article
Full-text available
Several approaches for formalising prerequisite structures on skills or competencies based on the psychological theory of knowledge space have been suggested and applied for adaptive eLearning. In this paper, we will discuss how these structures may be applied in skill management in a broader sense. After introducing some formal structures for prerequisite relationships between competencies, we will briefly present an example of an adaptive eLearning system based on this approach (APeLS). Several other aspects of the system which promise to be useful for advanced skill management are discussed. In the final part of this paper, we will discuss such broader applications of the model with respect to personal as well as to organisational skill management.
Conference Paper
Full-text available
Knowledge space theory was already applied for building adaptive testing and training systems which contain test and training problems including their prerequisite structure, and which offer explanations and help for solving these problems. Normally, however, lessons prepare students for solving problems. In this paper, the authors present a method for systematically structuring an adaptive eLearning course containing test problems as well as lessons. This method is based on knowledge space theory extended by component-wise representation of problems and on applying demand analysis. A course developed with such a method can be fed into an adaptive tutoring system, realized, e.g., within the adaptive tutoring system RATH.
Article
Full-text available
RATH is an adaptive tutoring WWW software prototype combining a mathematical model for the structure of hypertext with the theory of knowledge spaces from mathematical psychology. Using prerequisite relationships between different items in a domain of knowledge and using the knowledge about the pupil's current knowledge state RATH presents only those links in a hypertext document to the pupil which point to a document for which he/she fulfils all prerequisites and which he/she therefore should be able to understand. In a first prototype course, this idea is applied to the field of elementary probability theory. This tiny course is based on prior research of the second author. Thus, RATH is an important step in bringing psychological theories into a working tutorial system.
Article
The information regarding a particular field of knowledge is conceptualized as a large, specified set of questions (or problems). The knowledge state of an individual with respect to that domain is formalized as the subset of all the questions that this individual is capable of solving. A particularly appealing postulate on the family of all possible knowledge states is that it is closed under arbitrary unions. A family of sets satisfying this condition is called a knowledge space. Generalizing a theorem of Birkhoff on partial orders, we show that knowledge spaces are in a one-to-one correspondence with AND/OR graphs of a particular kind. Two types of economical representations of knowledge spaces are analysed: bases, and Hasse systems, a concept generalizing that of a Hasse diagram of a partial order. The structures analysed here provide the foundation for later work on algorithmic procedures for the assessment of knowledge.
Book
We have learned from Theorem 2.2.4 that any learning space is a knowledge space, that is, a knowledge structure closed under union. The ∪-closure property is critical for the following reason. Certain knowledge spaces, and in particular the finite ones, can be faithfully summarized by a subfamily of their states. To wit, any state of the knowledge space can be generated by forming the union of some states in the subfamily. When such a subfamily exists and is minimal for inclusion, it is unique and is called the ‘base’ of the knowledge space. In some cases, the base can be considerably smaller than the knowledge space, which results in a substantial economy of storage in a computer memory. The extreme case is the power set of a set of n elements, where the 2n knowledge states can be subsumed by the family of the n singleton sets. This property inspires most of this chapter, beginning with the basic concepts of ‘base’ and ‘atoms’ in Sections 3.4 to 3.6. Other features of knowledge spaces are also important, however, and are dealt with in this chapter.