Internationally there is a growing need for the development and cataloguing of training materials. At present, a large number of educational resources are available through various platforms, such as Wikis, GitHub and Moodle. However, such resources are not always easy to find and integrate into a learning course, amongst others, because the required metadata is not available. The reuse of educational resources is only possible when accurate and descriptive metadata is available (Roy et al 2010). The process of creating metadata can be very tedious and time consuming but is essential for sharing and reuse of educational resources because metadata allow teachers and learners to search or discover the most appropriate and well-suited educational resources possible.
The ENVRI-FAIR training catalogue has been designed and developed by following simple but effective requirements, mainly based on the need to find educational material. In order to identify the most suitable metadata set for the ENVRI-FAIR Training Catalogue, two well-known metadata schemas, namely Dublin Core (ISO 2009) and IEEE Standard for Learning Object Metadata (IEEE 2002) have been investigated. The Dublin Core Metadata Element Set (DCMES) contains 15 well defined elements for describing the "core" properties of digital and physical objects. However, it does not contain any elements that can be used for describing the pedagogical information of educational resources (Roy et al 2010). The IEEE Learning Object Model (LOM) consists of 60 optional elements that can be used to describe learning objects. Such elements can be combined in various manners to describe the pedagogical intent of an educational resource. This flexibility is important as the IEEE LOM can be too complex for novice catalogues. Mechanisms for converting between Dublin Core and IEEE LOM exist, but the aggregation and disaggregation of metadata fields can affect the usefulness of the resulting metadata. After reviewing the two schemas, a customized profile of IEEE LOM has been developed. Such profile consists of only 27 metadata elements, with respect to the 60 elements available within the IEEE LOM standard.