And Retrieval, And Ontology-based Geographic Information

GIS theory is becoming mature through more than 40 years and moves into a new developing phase. Computer can not understand data precisely because geographic information only describes data itself, the relationship between data is extremely rare. Ontology is a strong tool to model the sharing concept and academic circles has widely value its potential advantages in solving semantic heterogeneity, interoperability and semantic integration.This thesis is set in ontology and geographic information, based on ontology-based geographic information system. It centers on the establishment of geography ontology, the formal expression of geography ontology and its application in geography information query. This paper mainly researches in the following aspect:Firstly,The establishment of geography ontology. The process of establishing ontology includes establishing document, establishing classes and its hierarchy, establishing property and its ranges and adding examples. According to the geography feature, we use hierarchy to express geography concept and property. Special feature is a distinct aspect which geography ontology different from other ontology. We focus on Nature and social attribute.Secondly,The establishment of geography ontology case base. We combine the established geography ontology and any metadata which comes from data source to convert isomer to ontology with same data structure. Domain ontology can generalize basic feature of classes. Ontology examples can be established according to domain ontology's classes and properties. Finally,Query optimization and extension. We use query optimization and extension method to process searching key words to satisfied users'request. We optimize query by adding potential semantic word to the searching key words. We also use the hierarchy of classes (is-a relationship) or other relations between concepts and attributes as the foundation of query extension.The above mentioned research result shows that the precision ratio and recall ratio is proved by a big margin after query optimization compared to OGIS system.