Research on multi-agents information retrieval system based on intelligent evolution
Bin Xue, Yan Gui-ling · 2012
In view of some problems existing in current information retrieval system, this paper puts forward a multi-agents information retrieval system based on intelligent evolution, which is divided into nine modules: user Agent, communication Agent, mining Agent, personal Agent, intelligence evolution Agent, information retrieval Agent, group Agent, intelligence evolution filtering Agent and clustering Agent. According to the user history query information, current retrieval information and feedback information, this system rectifies the judgment of the user preferences constantly, makes the returned query results can reflect the user demand more and more, that is: provides users with the more and more accurate query result. Experimental results show that the proposed method provides the more accurate query on the document than the existing intelligent retrieval systems, the average of query accuracy is 81.6%, and along with the using system by users, and with the system running continually, query accuracy will be higher, which reflects the effect of the intelligent evolution.