The Research On Model Of Information Retrieval Based On Ontology And Multi-Agent

With the rapid development of the Information technology especially the application and extension of Internet,the information in Internet is massive which is growing geometrically.In such an infinite and chaotic space,it has already become a very difficult task to get information quickly and precisely that people need.How to search out user's required or interested information in such a vast Internet world has already become the target that researchers pursue.In order to achieve this target,the development orientation of searching technique should be close to user's requirement ceaselessly,simulate human's wisdom and the development with intelligence and personality has become an irreversible trend.In this paper,the Multi-agent system is in-depth studied and the conception of Ontology is introduced into MAS.Because the ontology can make semantic description of resources and the Multi-agent system has the characterisitics of intelligence and personality,with the combination of these two faces,we set up a simple Ontology instance in the computer field to standardize the user's requests, analyze the results returned by the search agent and study the user's interests,so it can look up information for users independently.The main jobs in this paper are as follows:(1) On the basis of analyzing the research status quo of the ontology theories, the searching engine technology and Multi-agent technology,a kind of information retrieval model based on Ontology and Multi-agent is presented in this paper.(2) Study the construction technology and the storage technology of Ontology thoroughly,then establish an Ontology instance in the computer field as well as make use of the Jean kit to let Ontology persistently into the database.(3) A kind of Agent communication language based on OWL is presented and the communication problems of Multi-agent in the information retrieval model are optimized by using the Ontology technology.(4) Each agent module in the model is designed and implemented by using the JADE development platform based on Java language and the model is tested with the computer hardware configuration data snatched from pages.Furthermore,compared with the traditional retrieval technology based on keywords,the experiment result shows that,this model of information retrieval based on ontology and Multi-Agent has better recall rate and precision rate in a certain extent.