Special issue: learning objects in context guest editors introduction.

Welcome to this special issue on "learning objects in context"! The papers in this issue have their origins in a workshop we organized in March 2005, with the support of the ProLearn network of excellence on professional learning (1) and the iClass project (2).

The papers in this issue focus specifically on the context in which learning objects are deployed: this theme is quite varied and picks up on the message that "if content is king, then context is queen." Overall, the main aim is to exploit data available about user context to:

* generate metadata automatically,

- either during the authoring process, from the author's context, when information about the intended use is more easily available; or

- while the learner is working with the learning object, in order to capture his feedback, which can then be used later to help guide decisions on the appropriateness of the same object for other learners;

* personalize the selection of relevant learning objects, by selecting objects that satisfy conditions that reflect the requirements of the context; and

* adapt the behavior of a learning object to the specific characteristics of the context.

Some of the recurring themes in the papers that follow include:

* Rich, "semantic" descriptions: A number of initiatives try to exploit upcoming new semantic web technologies in order to support more sophisticated descriptions. Indeed, this is one of the two opportunities evoked in the short paper by Baker. It is the essence of the SIMBAD project presented by Bouzeghoub, Defude, Duitama and Lecocq. Similarly, Heller, Steiner, Hockemeyer & Albert rely on domain ontologies to identify skills and their interrelationships.

* Adaptation strategies: It is clear that, in order to exploit context, some kind of adaptation is key, either in the selection of relevant objects, or by a particular such object itself. Heller, Steiner, Hockemeyer & Albert focus on how Knowledge Space Theory can be used as a framework for defining and organizing adaptation structures. Their approach models skills and competencies and considers the use of learning objects from a psychological theory that emphasizes knowledge assessment. In the paper by Specht & Kravcik, different adaptation strategies are discussed: their overview helps in putting their own work in the RAFT project into context (!).

* Pedagogical issues: O'Keeffe, Brady, Conlan & Wade discuss "pedagogically informed" adaptation strategies that they have developed as iClass services. Muehlenbrock focuses specifically on the problem of dividing learners in groups for a collaborative learning setting, and how that process can be informed by contextual information. The paper by Poldoja, Leinonen, Valjataga, Ellonen & Priha presents a specific kind of "progressive inquiry learning object templates" that support social constructivist approaches. Turker, Gorgun and Conlan focus on the pedagogical aspects involved with personalization, and how these are supported in the iClass approach.

* Small granularity of learning objects: By reducing the granularity, opportunities for adaptation to the context of use increase. Dahn details the slicing book approach that facilitates reuse of small text components in the domain of mathematics. The paper by Schluep, Bettoni & Schar presents a simple component model that defines "didactic content types": they have implemented this model in the dLCMS project, that also provides much appreciated guidelines to the authors. An alternative content model is the ALOCOM model presented in the paper by Verbert, Jovanovic, Duval, Gasevic & Meire: remarkable about this approach is that it enables automated dis-aggregation and re-aggregation of learning object components. The authors present a specific case study that works with Microsoft powerpoint documents.

* Modeling of context: One way to be explicit about the intended context of use is the definition of learning activities, which Sampson and Karampiperis consider to be the core of "next generation" learning systems. They discuss an early proposal for an architecture of such a system, and a toolset they have developed to experiment with it. Specht & Kravcik describe approaches for extending metadata with context information that can be captured automatically through sensors. Strijker & Collis detail several context dimensions that their research in a wide variety of learning settings has revealed to determine success of reuse strategies.

* Technical issues: Of course, context support needs to be integrated into the overall technical infrastructure. The paper by Massart discusses the merits of a recently finalized standard for querying repositories, in order to obtain not only the metadata, but also the actual objects themselves. Ternier & Duval report on experiences with the same standard for the implementation of access to a heterogeneous set of learning object repositories. Baker mentions the use of Service Oriented Architectures (SOA's) for modeling the processes that define the context. Paulsson & Naeve actually propose a SOA framework for learning environments that reflects their learning object taxonomy and that separates data, presentation and logic.

This collection of papers provides a balanced overview of some of the more recent and exciting work on context related issues for learning objects. We hope that you will get a useful overview of the context (!) in which some of this work is carried out, and, above all, that you may be able to assess the relevancy of what is rapidly evolving to your own work, either as a researcher or as a practitioner.

We are very grateful to the authors and participants at the workshop, which was a testimony of the dedication, energy and insight of this community. We would welcome reactions, comments and feedback from you, the reader.

Notes:

(1) http://www.prolearn-project.org/

(2) http://www.iclass.info/iclass01.asp

ERIK DUVAL, STEFAAN TERNIER

Dept. Computerwetenschappen, Katholieke Universiteit Leuven 

{erik.duval,stefaan.ternier}@cs.kuleuven.be

FRANS VAN ASSCHE

European Schoolnet

[e-mail]