The use of knowledge bases as repositories in learning environments.

Abstract

This dissertation proposes a method of management of Learning Objects Repositories (LORs). It looks at both the storage and search of learning objects (LOs) in the LORs. The focus is on how LORs can be structured in order to promote easier and more efficient search and thus furnishing users with relevant results. The model proposed defines a “neighbourhood”, a concept achieved by grouping LOs by purpose (genre-classification) using metadata elements from IEEE’s LOM. The elements chosen, which fall under the Educational and General groups, are Learning Resource Type and Keyword. This can then be put into a Knowledge

base to enable users to get additional results based on searches done by previous users who

requested similar objects. This involves the exploitation of similarities between users

(including the user conducting the search) to suggest additional LOs that could be relevant.

Two algorithms have been developed that use this model, one for searching and the other for insertion.