Metadata harvesting learning resources - an agent-oriented approach

The paper presents initial ideas and architecture for harvesting learning contents available on the web, with intention to support more qualitative usage and reusability of learning resources. Commonly-known web technologies are considered as good instruments to be used for harvesting and mining embedded metadata. Metadata harvesting/mining through different learning resources is by its nature distributed, and represents a good basis for employing agent technology. Based on previously implemented agent architectures at our university and extensive usage of e-learning tools we propose agent architecture for metadata harvesting in the context of learning resources.