Abstract:

Within the context of this paper, we understand Open Educational Resources (OERs) as freely available educational materials. In order to store OERs metadata in a structured form, modeling standards should be utilized. However, OER providers often do not employ such a standard. This leads to limitations in the development of federated systems that link multiple OER repositories to each other. Such federated systems, in turn, facilitate the search for OER across repository boundaries to help teachers, students, etc. to search for OER with different origins.

In this paper, we describe our methodology for connecting different OER repositories based upon the Learning Object Metadata standard (LOM). For this purpose, we have carefully selected 29 OER repositories that are openly available on the web and hosted by German institutions. Some of the key outcomes of our research are:
(i) only 17% of the analyzed repositories employ LOM as a standard to model their data,
(ii) only 12% of the metadata from the repositories will get lost when federating these repositories using the LOM standard,
(iii) the lost metadata has no impact on the quality of the retrieved OERs, as it is insignificant information about the OERs, such as the number of views and likes.

Based on this experience, we recommend that OER repositories use the LOM standard to describe OERs, as this has several advantages, such as facilitating the federation of repositories and increasing the accessibility of OERs.

This paper starts with motivation and puts our work in the context of other related research activities. The main focus is on the description of our methodology to federated existing OER repositories, which includes harvesting and mapping of metadata from different repositories. The paper closes with a discussion of our results and an outlook on future work.

Keywords:

Open Educational Resources, LOM, Open Educational Repositories, Federating Open Educational Repositories.