Multi-agent Model for Evaluation of Learning Objects from Repository Federations - ELO-index

Abstract

Availability and reusability are desired characteristics in order to guarantee the quality of Learning Objects (LO) and, because of that, the implementation of metrics for these characteristics is important for their evaluation. This paper describes an approach that uses a Multi-Agent System for assessing the LO, applying different methods and metrics and finally weighing them to obtain an index called ELO-index. Using metadata as our source of information, the metrics used for calculating ELO-index was completeness, consistency and coherency. The obtained index can be used to recommend LO by matching them with user-provided keywords, but also to manage the repository in which they are stored evaluating their quality before being published.