Repositories form a central piece of the learning objects technology, providing storage spaces where the objects can be catalogued, located, and retrieved. However, repositories usually support only syntactical and morphological aspects of learning objects metadata for cataloguing and searching purposes. This article proposes two integrated systems, which provide the core services of a semantic repository of learning objects. MSSearch system uses ontology alignment techniques to create a semantic search engine and a semantic database for learning objects metadata. MSSearch supports the integration of multiple educational ontologies through a combination of ontology aligning and mapping mechanisms. In turn, Linnaeus system offers intelligent support for the creation and editing of learning object metadata. This tool employs the technologies of intelligent agents and educational ontologies to provide an intelligent semi-automatic metadata filling service. This article presents the main architectural components, ontological models and interface facilities of these systems. The text finishes with the presentation of the experiments conducted to validate Linnaeus and MSSearch.