Design And Realization Of A Information Retrieval System Based On Domain Ontology

The information retrieval system based on keywords establishs the keywords of the original information inverted index. When users enter a query keywords, the information retrieval system search the keywords the user enters from the inverted index library,then documents that contain keywords. The information retrieval system based on keywords only match the keywords from the the inverted index library,the retrieval recall and precision rates are relatively low.So,now the semantic retrieval system in some field emerge.the semantic retrieval system is based on domain ontology.the system employs the domain ontology label the original text information the RDF(Resource Description Framework) data format. When a user search something, the semantic retrieval system convert the keywords the user enter RDF format,and then match it in the RDF database. Currently,the way of the RDF data storage includes relational databases, Sesame, RDFDB and Redland.This paper realize the the inverted index structure,and test its retrieval efficiency. employ B+ tree structure store RDF data,and employ experiment compare the retrieval efficiency with storing RDF data in relational database and in RDF data files. realize a information retrieval system based on keywords,build a small ontology about maize,and employ this ontology extend the information retrieval system.Through experiments,this paper find these result.First,the inverted index structure suit the information retrieval system.Second,The time search RDF data storing in B+ tree is much smaller than that storing in the RDF files,and a little smaller than that storing in the relational database.Last, employing the ontology extend the information retrieval system can increase the recall and also can be used for navigation,the ontology guide the user to retrieve the information.