Automatic learning object extraction and classification in heterogeneous environments

Abstract

This paper proposes the use of federated databases techniques in searching for educational resources by using a learning object paradigm that describes these resources based on metadata. Combining a complete agent-based architecture that implements the concept of federated search along with IR technologies may help organizing and sorting search results in a meaningful way for educational content. The paper presents also the ground for an approach for semantic-aware learning content retrieval based on abstraction layers between the repositories and the search clients. © 2011 Springer-Verlag Berlin Heidelberg.