Marek Meyer

Multigranularity reuse of learning resources

This article investigates a scenario of reuse in which existing learning resources serve as preliminary products for the creation of new learning resources. Authors should be able to reuse learning resources and also parts of them at different levels of granularity in a modular way. The requirements of multigranularity reuse are analyzed and compared to existing solutions. A concept for modular, multigranularity reuse is presented in this article. It is also shown how this kind of reuse can be achieved in practise.

Requirements and an architecture for a multimedia content re-purposing framework

Production of Learning Resources is very expensive. Hence re-use is a common way to lower these costs. In contrary to a re-use ”as is” re-purposing means to transform a Learning Resource to suit a new learning or teaching context. We consider re-purposing to be a promising approach to improve the re-use of Learning Resources. As a basis for re-purposing tools, a re-purposing framework shall provide fundamental functionality in order to facilitate the development of re-purposing tools. This paper analyzes the requirements for such a framework, proposes an actual architecture and presents the current state of the implementation.

Automatic classification of didactic functions of e-learning resources

Re-use of digital resources is an important issue in e-Learning scenarios, because only intensive re-use can make e-Learning cost efficient. Besides reusing whole courses, authors often desire to re-use fine grained parts of courses for creating new Learning Resources. The granularity which appears to be most promising for this kind of re-use is the level of information objects. Information objects each have a dedicated didactic function; a set of information objects with different didactic functions are combined into Learning Objects. This paper analyzes how didactic functions of existing information objects can be automatically classified using machine learning technology. The results of such classification methods on a set of Learning Resources from medical science are discussed.

Using community-generated contents as a substitute corpus for metadata generation

Metadata is crucial for reuse of Learning Resources. However, in the area of e-Learning, suitable training corpora for automatic classification methods are hardly available. This paper proposes the use of community-generated substitute corpora for classification methods. As an example for such a substitute corpus, the free online Encyclopaedia Wikipedia is used as a training corpus for domain-independent classification and keyword extraction of Learning Resources.

Modeling modifications of multimedia learning resources using ontology-based representations

Repurposing of multimedia-based Learning Resources is an important issue in E-Learning, as economic success of content production depends on how intensively content is used. Repurposing does not only mean reuse ”as is”, but also comprises modifications of the contents to suit a different learning or teaching context, as well as reuse of fragments of a large Learning Resource. This paper introduces a method for modeling multimedia content modifications based on an ontology-based content representation. A theoretical background for modeling modifications of multimedia contents independent of the particular format is provided. Also a practical implementation is presented and discussed.

A concept for cross-layer optimization of wireless sensor networks in the logistics domain by exploiting business knowledge

Energy is limited in wireless sensor networks due to mostly battery-operated wireless sensor nodes. Consequently, an efficient energy usage significantly enhances the lifetime of a wireless sensor network. In this paper, we propose a cross-layer optimization concept that exploits business knowledge in the logistics domain to adapt communication by business relevancy. A corresponding information value is assessed for each detected event. Energy and communication costs can be saved by dropping or postponing the transmission of less relevant event information, without negative impact on the business and application level.

Modularization and multi-granularity reuse of learning resources

This thesis investigates modular reuse of learning resources. In particular, it considers a scenario of reuse in which existing learning resources serve as preliminary products for the creation of new learning resources for Web based training. Authors are interested in reusing the learning resources created by other authors. It is assumed that these authors belong to different organizations. Furthermore, these authors do not use a common authoring tool because they are obliged to use the tools specified by their respective organizations.

Improving authoring-by-aggregation and using aggregation context for query expansion

Authoring-by-Aggregation is an authoring paradigm for creating new Learning Resources by composing several smaller ones. However finding suitable Learning Resources that can be reused in a particular course is a time-consuming task. The user has to query a repository and review the resulting list of Learning Resources whether they are applicable. In this paper we present a new method for narrowing the result set by taking the aggregation context into account. Context features are used as additional query attributes, leading to more precise query results. This paper also presents an improved Authoring-by-Aggregation process and a corresponding implementation, which facilitates the creation of new Learning Resources.