A Meta-Relational Approach for the Definition and Management of Hybrid Learning Objects

Electronic learning objects (LOs) are commonly conceived of as digital units of information used for teaching and learning. To facilitate their classification for pedagogical planning and retrieval purposes, LOs are complemented with metadata (e.g., the author). These metadata are usually restricted by a set of predetermined tags to which the classification schema must conform (e.g., IEEE LOM). In our experience, certain complex LOs need to be complemented with different types of domain-dependent information for their pedagogical planning and retrieval: (i) classification metadata for enhancing contextualisation, search and retrieval (e.g., the tagged structure of an archaeological site where an archaeological object has been found) and (ii) additional data that can enrich the LO (e.g., the weight and other dimensions of an archaeological artifact described in a podcast). We refer to LOs enhanced with domain-dependent information as hybrid learning objects (HLOs). However, most learning object repositories (LORs) only permit a predetermined, fixed set of metadata attributes to be used in the classification of LOs. This rigidity is inappropriate when domain-dependent information schemas are used for the browsing, retrieval, and domain-specific pedagogical sequencing of HLOs. Thus, custom software applications are needed to manage LOs that must be tagged with information belonging to specific domains. This paper presents a theoretical approach that permits the use of a single LOR for classifying and enriching LOs according to domain-dependent information schemas, which can be dynamically changed after their definition. The key issue in our approach is the presence of a meta-relational model for the dynamic definition of specific domain-dependent relational database schemas used for classifying and enriching LOs.