Abstract:

These days education related institutions like universities need e-Learning systems, which help to distribute learning content and to share information and assessments within the community. Currently systems from different companies or communities represent the e-Learning environment in most of the education institutions. Learning Management System (LMS) for representing the content of lectures and exercises are mostly the base for e-Learning. Media systems for streaming video data, e-assessment tools and other tools are creating additionally a heterogeneous e-Learning environment. This leads to challenges for the learner to understand the use cases and the workflow of the different systems and to find all accessible content. A heterogeneous environment normally cumbers the learner and creates additional expenses to access material in different systems. Moreover it is difficult for a teacher to use and administrate all systems. The creation and maintenance of an up to date knowledge collection for the learner is a time-consuming task for the teacher.

We present a concept which represents knowledge in a separated repository by a common structure. The content is organized as a graph and contains links to different systems. Each node in this graph represents a topic, like mathematics, psychology or software development. Each topic can have a set of links to different systems for education. Edges between the nodes represent the dependencies between the topics. The topic computer science for example can have connections to software development, programming languages and project management. This kind of knowledge representation helps the learner to access knowledge cross-system wide, which enables effective learning. Important is a usable and clear graphical representation of the learning content and connection among the topics. When new e-Learning content is created in a system, keywords can be defined by the author, which represent the content for our knowledge base.

These keywords need to be mapped into our multilingual graph, consisting of topics and their dependencies. To realize this mapping a defined set of meta data attributes need to be transferred to the knowledge base. For example the structure of the education institution where the current element is located can be used to map the course into the right position of the knowledge base. Existing content in e-Learning systems can be mapped semi-automatically by wizards and author-defined keywords to elements into the knowledge representation. Different types of knowledge links can be defined. For example a link which represents literature, a link into a resource in a LMS or a link to an assessment form. The presented concept and its implementation are planned as standalone system which can be used by all LMS or e-Learning related systems. The first realization is planned in the area of the LMS OPAL, which is used by Saxon universities in Germany. OPAL and other e-Learning tools are developed mainly by the company BPS GmbH with support of different education institutions.

Keywords:

Cross-System Knowledge Representation, LMS, Multilingualism.