Olympus: Personal knowledge recommendation using agents, ontologies and web mining

Abstract

There are many initiatives in the scientific community to produce knowledge management and CSCW systems. However, it is difficult to promote the easy information share among learners. In this paper we present Olympus, a multi-agent system to help learners share not only what the information content is, where the information is, and who have the information needed by the learner, but also how to use the available knowledge. Olympus uses agent technologies, ontologies and data mining to create knowledge chains in a semi-automatic way, which is a job mat usually would take a lot of effort. The agent monitors the learner's web navigation activities. From there, another agent classifies its content using an ontology, creates and recommends a knowledge chain to the learner. As a sub-product of this work we establish a knowledge base with classified web pages contents. © Springer-Verlag Berlin Heidelberg 2007.