Interfaces with Other DisciplinesUsing similarity measures for collaborating groups formation: A model for distance learning environments
Introduction
More than sixty years after the introduction of the first computer system, Internet has become the standard platform for e-learning environments. E-learning, the contemporary version of distance education, is mainly web-based, conducted by means of Internet-connected computers running special programs (learning content management systems, LCMS), which bring learners, teachers, courses and collaborative technologies into contact.
Adaptive e-learning systems, an alternative to the traditional approach in the development of e-learning platforms, use a variety of methods in order to adapt to the needs of each separate user. Modern, advanced information and communication technology has to be used in more ways than simply retrieving learning material. Research on adaptive e-learning is almost ten years old, yet adaptive learning environments are mainly research prototypes with little, if any, standards compliance (Paramythis and Loidl-Reisinger, 2004). There are two major drawbacks obstructing broad use (Brusilovsky, 2004): lack of integration and lack of re-use support. However, both drawbacks may be resolved successfully by the use of learning objects (LO) technology (Alvarado-Boyd, 2003).
A vast variety of definitions of LO can be found in literature. In a recently published paper a Learning Object is defined as a “standalone, reusable, digital resource that aims at teaching one or more instructional objectives or concepts” (Mavrommatis, 2006a). The Learning Objects Metadata (information used to describe a LO) framework is described in the Sharable Content Object Reference Model (SCORM, 2004) and several learning objects repositories (pools containing retrievable LO) are already in common practice for distance learning.
Brusilovsky (1998) presents a review of adaptation technologies; among them, adaptive collaboration support is defined as the technology that uses system’s knowledge about different users to form a matching collaborating group. Most current web-based educational systems collect large amounts of information about the students but this information, so far, is not widely used by instructors (Mazza and Dimitrova, 2004). Supporting collaborative learning is one of the most recent approaches of adaptive educational systems. The use of collaborative methods can extend e-learning from an individual learner to a group of learners (Mödritscher et al., 2004).
Collaborative learning activities are based on constructivist learning theory (Wilson, 1996) and, although collaboration in the classroom has proven itself a successful learning method, online collaborative learners do not seem to enjoy the same benefits, mainly because distance learning technologies do not provide guidance nor direction during online discussion sessions (Soller and Lesgold, 2003). Most of the commercial, standards-based e-learning platforms currently used in higher education institution, allow very little collaboration by simply providing basic tools (Van Rosmalen et al., 2004). To make things even worse, Fung and Yeung (2000) found in literature fifteen research adaptive educational systems that were then reviewed to check their level of adaptivity. They were found to support a subset of the known adaptation technologies. Among them however none was reported to support adaptive collaboration.
On the other hand, the importance of collaboration is increasingly underlined by researchers and learning theorists: cooperative learning, communities of learners, social negotiation etc, are some examples (Wiley, 2003). Collaboration occurs when learners somehow work together to accomplish shared learning goals (Johnson et al., 2000).
In order to achieve maximum benefits, collaboration has to rely on well adjusted learning teams, therefore placing users (learners) randomly in a group and assigning them a task is not enough (Soller, 2001). Interacting with other people is crucial for a contemporary learning environment, and “interactive” learning (Tapscott, 1998) requires adaptive learning, which embodies adaptive collaboration. The first step in directing collaborative learning environments is, therefore, forming the right group(s) of learners. Additionally, re-use of a collaborative environment on a variety of courses is essential; therefore, it is necessary that adaptation tasks will be domain independent (Pollalis, 1996, Gaudioso and Boticario, 2003).
This paper presents a method for course construction that promotes collaboration in order to achieve a common educational objective for a community of learners. When doing the group work it is useful to form some subgroups bearing in mind common (or differentiating) student aspects. The method creates properly matching collaborating groups and at the same time selects appropriate learning objects to form the corresponding course’s core for each group.
The paper is organized as follows: the following section contains a mathematical model, based on simple information vector spaces, where we present both Learners and Learning Objects as the basis for similarity coefficients within educational technology; in Section 3 the Educational Cells are formed by applying clustering methods to the Learning Objects–Learners array; an application-example is presented in Section 4; in chapter 5 a few additional options are presented, aiming to integrate the model within a modern distance learning environment; and, finally, some conclusions are drawn in Section 6.
Section snippets
Resemblance coefficients in learning technology
Task Analysis, probably the most important component of Instructional Design, includes methods like Learning Hierarchy Analysis, Learning Contingency Analysis (Jonassen et al., 1999), or even, Principled Skill Decomposition (van Merriënboer, 1997). In general, these methods presume that every knowledge field or complex cognitive skill to be taught can be broken down into constituent skills, finally leading to construction of a learning hierarchy (similar to an ontology). A detailed description
Parametric clustering for group formation
The idea that (demand-driven) production systems share many similarities with educational systems, is more than ten years old (Kester et al., 2001). By analogy to the machines-parts table in Manufacturing Cell Formation, an application of Group Technology to production, we define the Learning Objects–Learners table which we call the learning compatibility matrix (LCM).where ∂R, ∂GR ∈ (0, 1].
Creation of Educational Cells (groups of
A learning community groups its members towards a common objective
As an example, let us assume a certain knowledge field, where Task Analysis yielded a 5-dimensional Information Space S = {s1, s2, s3, s4, s5}. The Repository contains t = 10 Learning Objects and the Community has k = 6 Learners. All Learning Objects’ and Learners’ properties vectors, together with computed corresponding General Resemblance (GR) and Relevance (R) coefficients between each couple, are shown in Table 2.
Algorithm Educational-Cells produces several consistent LCM instances. Instances with ∂R ∈
Further applicability issues: towards the formation of most valuable distance learning environments
In the preceding sections we have presented a method that outputs groups of learners with similar educational background, together with a corresponding set of Learning Objects for each group. These sets are serving each group as their Course’s Core (CC) towards the overall common objective (Fig. 2).
The method takes into account the learners’ knowledge and the available learning objects content, and can be easily automated, thus providing support and advice to tutors and learners. Yet, there are
Conclusions and future research
Distance learning technologies and e-learning methods seem to have a common denominator: using interconnected information systems, which through special programs and platforms bring users/learners in contact with courses and teachers (LCMS). E-learning technologies are already in relatively mature stage, while research in the area advances in what lately is known as Learning Objects.
This paper’s main goal is to automate learning groups and environments by matching educational means and
References (62)
- et al.
Just-in-time information presentation and the acquisition of complex cognitive skills
Computers in Human Behavior
(2001) Knowledge representation and construction in hypermedia environments
Telematics and Informatics
(1996)The resemblance coefficients in group technology: A survey and comparative study of relational metrics
Computers & Industrial Engineering
(1996)Hierarchical task analysis: Developments, applications, and extensions
Applied Ergonomics
(2006)- et al.
Similarity coefficient methods applied to the cell formation problem: a comparative investigation
Computers & Industrial Engineering
(2005) Traveler’s Guide to the Learning Object Landscape
(2003)- et al.
The Jigsaw Classroom
(1997) - Bekele, R., 2005. Computer-Assisted Learner Group Formation Based on Personality Traits. Ph.D. dissertation, University...
- et al.
Future directions of blended learning in higher education and workplace learning settings
- et al.
A new reconfigurable test vector generator for built-in self-test applications
Journal of Electronic Testing: Theory and Applications
(1996)