| Currently,the computer is changing from the sole equipment to the worldwide scale network of the information exchange and business process.Therefore, the duty which computure must face urgently now is exchanging, reusing, and sharing of data, information and knowledge. So it is necessary to research and develope Ontology in knowledge engineering, natural language processing, information retrieval system, intelligent information integration, knowledge management, exchanging of information and software engineering fields and so on.The method of gaining Ontology is not mature. At present the knowledge establishment in Ontology library is carried on manually with the participation of the domain experts. This is a very difficult task and manual operation can not guarantee the completeness of Ontology library.Ontology should be gained automatically through the knowledge excavation.It is also one of the Ontology researchers' primary missions.The glossary semantic classification on this paper is a part of a software about the Ontology to produce automatically and simultaneously may be applied in the automatic translator technology area .A new mentality about glossary semantic classification is given and a system about glossary classification based on semantic is designed. This method must first carry on the segmentation to the dictionary or the article and obtain the glossary correlation data. Word vector and glossary semantics similarity data can be obtained through data processing. Then the glossary hierarchy tree is constructed according to the glossary relation about hypernym and hyponym. The optimizing process is carried out to optimize the glossary semantics classification effect, reflecting the classification relation better from the semantic angle. Finally, the hierarchy relation of glossary semantic classification is shown on web page for visualization. It turns out that the semantic classification of glossary can be realized with this method. |
PDF Full Text Request