The Information Retrieval Based On Geo-Ontology

With the rapid development of information technology, network has become a main means of obtaining information. It brings convenience to people, at the same time, also brings some troubles. As information in network is very large, people often spend a lot of time to search for useful information. When people travel abroad, they aren't familiar to food, drink, housing, and walking in an unfamiliar urban environment. So some geographical information is retrieved, such as bus stations, bus lines inquiry and so on. Relative trip guidance is obtained by network for consulting useful information.Because currently search engine is basically based keywords precise matching retrieval algorithms, the query results are redundant and useful information can not be clearly found. This is clearly unsatisfactory, and user's loyalty and satisfaction about search engine will be greatly reduced. Ontology, as a new method of semantic model, is mainly used for knowledge sharing, interaction and reuse. Ontology has a good concept of hierarchy and logical reasoning. Semantic retrieval which is based on the ontology is achieved by the concept of ontology construction levels, which establishes the concept and the concept of relationships between attributes by using ontology, and finds uncertainty, indirect information implicit in between concept, thus can realize the semantic intelligent information retrieval.In this thesis, current status and deficiencies of information retrieval are analyzed firstly, and ontology related technologies are introduced, the basic concepts of ontology, description language and construction methods are explained. Ontology of geographical information in domain ontology is built by the tool of protege. The specificity of geographic ontology is described, meanwhile the special space features in geographical ontology are expressed by OWL language.At the second place, several key kinds of spatial relations are described, fuzzy sets are introduced into distance relations in spatial relations, and then the membership functions about distance concept constructed, which can express the concepts of "near" and "far" that make queries more comprehensive. Third, spatial direction relations are mainly studied, query expansion about them is carried out, and the algorithm's results are demonstrated by experiments.Finally, the algorithm is applied to actual research. According to the local public traffic, the corresponding transportation ontology is established which can search bus stations of a region or a direction intelligently, and then the search algorithm is improved further in order to verify the feasibility and practicality of the method.