Abstract

Due to the development of the World Wide Web, in the number of contents, nodes and transmission speeds, the scientific community is making proposals for different uses. These uses are nowadays far more ambitious than downloading files to be rendered to the user.

The Semantic Web is developed with this aim. The Semantic Web includes information which can be understood by Intelligent Agents (IA). In this process, the IAs can expand their Knowledge Bases (KB) and perform inferences that can ease some human tasks. The information in the Semantic Web uses different ontology languages for the web, as the World Wide Web Consortium (W3C) recommendation named Web Ontology Language (OWL). The inferences can be done by different Inference Engines (IEs) as FACT++, Racer and Pellet.

In addition to this, many educative contents have appeared in the web, with high production cost. For this reason reusing contents is nowadays a need, with little or no changes. These units of contents to be independently shared are know as Learning Objects (LOs). These LOs have to interoperate among them in order to build e-learning experiences.

The document presented fosters this LOs interopeability. To achieve this objective, a complete ontology is specified, based on Description Logic (DL) and developed in OWL, in order to be used in the Semantic Web. In this ontology, concepts related to the different relationships among the stakeholders of an e-learning experience are studied, developed and implemented.

A new architecture called Semantic DHT is defined in order to exploit the ontology. The Semantic DHT is a Peer to Peer (P2P) architecture based on Distributed HastTables (DHTs), and allows the Intelligent Agents access to the ontological information, with a high node fail tolerance. These Agents facilitate the ontological searches done by the users. Finally, the ontology and the architecture are validated against real e-learning experiences.