Research On Domain Semantic Retrieval Model Based On Ontology

In the information age we wouldn’t worry about the lack of information, but in the mass of find information accurately and quickly, improve the efficiency and accuracy of information retrieval. Let retrieval with semanticc, achieve good man-machine communication and make computer understand human needs. Semantic retrieval provides an effective and reliable method f or the large amount of data for retrieval. Ontology within ascertain range of information by defining the way to organize, share information and semantic information in the semantic retrieval is realized, ontology query expansion based on semantic relations, calculation of matching the search results by using the inference rule of semantic similarity, semantic reasoning retrieval results, the preferred display search results the front of the sort.Firstly, this paper uses the ontology construction method build ontology,it can use a standard way to organize the field information, determine the concept and the semantic relationship, also design the inference rules.Specifically improved calculation method of similarity base on ontology, analysis influence factors, established the calculation model of concept semantic similarity based on the candidate set. Finally, design semantic retrieval prototype system with constructed ontology and improved similarity calculation method. The system is composed of query expansion, semantic reasoning and retrieval display. Using precision and recall rate to verify the system effectiveness, the experimental results show that the semantic retrieval not only better than the keyword retrieval and also can meet the requirements of u sers. It provides a feasible scheme for constructthe actual semantic retrieval system.