Relation-based heuristic diffusion framework for LOM generation

Abstract

Learning Object Metadata (LOM) intends to facilitate the retrieval and reuse of learning material. However, the fastidious task of authoring them limits their use. Motivated by this issue, we introduce an original method for LOM generation based on relations between LOM documents. These relations significantly influence the attribute values. We formulate this influence with heuristics of acquisition, suggestion and restriction. A diffusion framework for these heuristics is suggested. In the context of relation-based graphs of LOM documents, this framework models the recursive processing of the heuristics. The generated values could then be used to assist users in generating LOM documents