| Since the concept of semantic web was proposed by T.Berners-Lee in 2001, more and more scholars have researched on it. Compared with the WWW, the semantic web can do better at realizing "careful, exact, automatic" search about the information. How to divide ontology is closely related to realize automatic search of semantic web information. The traditional ontology mainly uses the semantic relations of crisp concepts to query information. According to uncertain characteristic of natural language, it's necessary that fuzzy concepts and fuzzy relations are applied to ontology. Along with the society progress and development, it's true that employment of students is essential to the overall development of country, however, most of recruitment conditions of companies are fuzzy conditions, the semantic query between fuzzy concepts cant's be performed in the traditional ontology, it can't check overall and exactly. In order to handle fuzzy and uncertainty of employment information, a fuzzy ontology model which is capable of fuzzy concepts semantic query will be established, it is an important method to resolve the problem. So research on how to divide fuzzy ontology for students information domain is of great importance.In this paper, we firstly present a four-layer fuzzy ontology based on fuzzy domain ontology and fuzzy top-level ontology, then present basic fuzzy ontology. Secondly, aiming at the domain of students information which included nature situation, rewards situation and so on, the knowledge-structure of students information is expressed as ontology. We establish the four-layer fuzzy ontology for students information domain. Finally, the four-layer fuzzy ontology model will be applied to employment of students. Based on the recruitment conditions of companies, we calculate students' comprehensive evaluating by summing and weighting, and then semantic query between fuzzy concepts are established by computing with words inferencing. Then the four-layer fuzzy ontology accomplishes the searching of information and the condition of omitting information of coincidence semantic, and achieve the purpose of improving the overall and exactly search of information. |
