A similarity-based approach to enhance learning objects management systems

The Learning Object Repository is an important element in the management, publishing, location and retrieval of Learning Objects. However in this system, the task of finding Learning Objects that conform to user's needs and requirements still presents problems related to easiness, simplicity and usefulness in selecting and recovering them. Thus, this paper proposes a metric to determine the similarity among Learning Objects based on information represented in their metadata. The proposed method is based on a fuzzy approach to determine the semantic similarity of XML structures used to represent the Learning Object metadata. The metric has been implemented into a Learning Object Management System, as part of the Searching component. This component offers several searching mechanisms based on Semantic Query, Learning Objects Patterns and Query by Examples. The first results confirm the effectiveness of the approach in identifying similar Learning Objects, an important task in managing these educational resource types.