A Methodology and Framework for the Semi-automatic Assembly of Learning Objects
Katrien Verbert1, David Wiley2 and Erik Duval1
1
Dept. Computerwetenschappen, Katholieke Universiteit Leuven, Celestijnenlaan 200A, B-3001 Leuven, Belgium {Katrien.Verbert, Erik.Duval}@cs.kuleuven.be
2
Instructional Design and Technology Department,
Brigham Young University, Provo, UT, USA David.Wiley@byu.edu

Abstract. One of the major obstacles in developing high quality content for learning is the substantial development cost and effort. In addition, the return on investment is often low, as developed learning materials are difficult to reuse and adapt to new and different educational contexts. In this paper, we present a semi-automatic content assembly methodology to automate, at least partially, the reuse of existing learning content in high quality and effective learning sequences. In addition, we present a case study that integrates the approach into the LAMS learning design environment. Keywords: learning object reuse, learning object metadata, learning design
1 Introduction
Many existing course documents merge the representation of content and the instructional approach [1]. Such hardwired pedagogy restricts the options for teaching and learning, both in terms of reusability and adaptation of learning sequences. Typically, teachers create their teaching strategies and content from scratch or reuse parts of existing course documents by ad-hoc and time-consuming copy-and-paste actions. In addition, adaptation to individual learning or teaching styles, background, experiences, interests or preferences is generally not possible, unless learning content is specifically designed for personalization purposes [2]. In this paper, we present a semi-automatic content assembly methodology for the generation of learning sequences tailored to different pedagogical approaches, based on the explicit design of these sequences by a teacher. The assembly framework employs a teacher model, an instructional model and a domain model to enable the focused retrieval and aggregation of learning resources into learning sequences. Learning resources are retrieved through the GLOBE network of educational repositories [http://globe-info.org/] and from various community driven websites, such as WikiAnswers.com, ProProfs.com and Wikipedia. The assembly framework is

described in the next Section. We present a case study that integrates the approach into the LAMS [3] learning design environment in Section 3. Related work is discussed in Section 4, followed by conclusions and remarks on future work.
2 Content Assembly Framework The content assembly framework supports the selection and assembly of existing learning resources. The framework employs the following models to enable the focused retrieval and aggregation of resources: - The instructional model captures the semantics of the pedagogical strategy employed by a learning sequence and is based on [4]. Narrative structures within this model outline the flow of concepts of a particular learning design strategy and are used as templates when assembling learning sequences. An example is an inquiry based learning strategy that sequences activities like
"answer questions", "vote on a list", "discuss responses", "read expert view",
"discuss expert view" and "personal reflection". - The domain model represents the knowledge domain of a course. It includes concepts outlined in the objectives of a course and their interrelationships. - The teacher model defines teacher attributes to enable the personalized aggregation of learning resources [5]. The model includes attributes for representing the level of expertise of the teacher, interests and activities, teaching strategy preferences, background, and presentation styles (Fig. 1).
Fig. 1. Semi-automatic content assembly framework