A Multi-Agent Information Retrieval System Based on Ontology

Traditional information retrieval systems are lack of semantic comprehension, and also have inherent ambiguity of short keyword queries. To solve this problem, this paper proposed a Multi-Agent Information Retrieval System based on Ontology. Introducing ontology to information retrieval system can realize knowledge domain-expression in order to provide users with the increment of information service and refine the initial query. Furthermore, group agent use improved collaborative information filtering algorithm to construct the group preference base equal to the compilations of all the term preference sets of the similar users served by personal agent, so as to accomplish the personalized information retrieval according to users' interest of the group preference base and rank the retrieval results by comprehensively considering of the users' preference about a certain theme and the frequency of the keywords of the retrieved document that appear in the relevant domain.