Research Of Semantic Retrieval Based On The Ontology

A new intelligent information retrieval techonolgy is need to research urgently to loose the crisis of"rich data poor information"and use the information resource in Web better. Semantic Retrieval as one of two big developing direction of intelligent information retrieve techonolgy, occupies extremely important position in Intelligent Information Retrieval field. Semantic Retrieve based on the Ontology that this paper researched is the Semantic Retrieve based on the concept made by the Ontology, which not only have some semantic processing capacity and good NLI, but also can process the problem of the relation between the related concepts. So Semantic Retrieve based on the Ontology has important research value.First,this paper proposed a new model system framework of the semantic retrieve based on the Ontology. This framework has many advantages such as high performance, deploying simply, upgrading and making mantentant conveniently, etc. Second,this paper proposed a new method to build domain ontology which is named Object Oriented Building Method. This method owns many metrits, such as being simple to be learned and used, can decrease the workload in the process of developing the system of the Semantic Retrieval. This method affords a new idea to resolve the problem of building the ontology automatically. Third, this paper implemented one system of the semantic retrieve based on the Ontology, which is Fruit Semantic Retrieve System (FruitSRetrieve) in agriculture field, which proof the feasibility of the framework and Object Oriented Building Method. It is universal that the difficulties are resolved in implementing this system, so it has important reference value in the process of implementing semantic retrieve system in other fields.There exist many unsolved problems in Semantic Retrieval based on the Ontology, such as ontology formalization being incomplete, the evaluation of the ontology lacking uniform standard, the increasing information not being updated, etc. but with the more and more people with lofty ideals paying attentation to and researching the knowledge of the Information Retrieval and Ontology, these problems must be resolved or modified at last.