Research On The Ontology Construction And Its Application In Pension Insurance Domain

With the wide application of ontology in fields of semantic web, electronic governmentaffairs, data mining, and with the potential application in some new fields, such as cloudcomputing, physical networking, the application of ontology has drawn the attention of lots ofresearchers. As the foundation of the ontology application, the construction of domainontology has become the most urgent task. Due to the complicated characteristics of thedomain knowledge, there is no unified standard for the construction of ontology. Although theeffect of ontologies by the artificial building way is better, the amount of work about theontologies building will lead to longer construction life cycle.Under the circumstance above, in this thesis, many methods for ontology constructionand visualization are studied, and an approach for semi-automatic construction method baseon Pension Insurance is presented. In the approach, kernel ontology is given by domainexperts firstly, and then an algorithm based on statistics and rules is used to extract domaintaxonomic concept and non-taxonomic concept from social security lexicon. Initial ontologyis extended continually based on adding algorithm under the help of ontology visualizationand domain experts. Finally, based on the Pension Insurance ontology which we haveconstructed by above method, an application research of ontology on text clustering analysisis presented under the semantic characteristics of domain ontology. Though the experimentalanalysis, a good effect is presented. This proved the correctness of the methods which arepresented in this thesis.