Framework of multi-agent information retrieval system based on ontology and its application

In order to improve performance of existing information retrieval technology based on keyword matching, this paper introduces a framework of multi-agent information retrieval system based on ontology according to ontology theory and multi-agent theory. The design thought of system, functions of every agent, circulating mechanism and entire searching process are described in detail. This system model is applied in the urgency management of communal burst event after domain ontology is established under the collaboration of domain experts with ontology constructing engineers, can effectively makes information retrieval ascend to semantic level from syntax level, save user's searching cost, and increase searching efficiency.