Research On The Search Technology Of Geographical Information Based On Semantic Similarity

With the rapid development of Internet technology and rapid growth of the amount of information on the network, it is harder and harder for us to find the information that we exactly want.For about 80 percent of human information is spatial information, the diversity, distribution, heterogeneous characteristics make it much inconvenience of the spatial information sharing and interoperability. Traditional keyword-based information query retrieval mechanism, by not "understand" the semantic information which led to the serious shortageof precision and recall rate, which makes the current network search methods can not meet the increasing information retrieval needs.the emergence of Geo-Ontology is a good solution to this problem, which is the knowledge or entity in the field of Geography formed by a certain system of relations, so that the semantic representation of geospatial information and semantic sharing can be achieved. Therefore, this paper relies on 863 national project in the Intelligent Spatial Information Services section, the semantics of geographic feature ontology of geographic information search-related technologies research.This paper discusses the related theory and characteristics of information retrieval technology and analysis of it'sl problems and deficiencies. To deal with these problems, ontology-based semantic similarity is proposed. In the absence of reusable Geo-ontology, we research the techniques of the mapping from the geodatabase to Geo-ontology. In order to improve the Ontology's efficiency, we construct the geo-ontology reasoning rules to expand geo-ontology,and dig out the explicit knowledge. Also, based on the achievement of semantic similarity, the semantic similarity calculation model was proposed to improve the existing models, take advantage of the good level of semantic relations of ontology to transform the key-based search method to semantic-based method.Based on the above theoretical study, the geographic information search engine is designed, and achieved semantic similarity search function based on geo-information. In order to facilitate users to search, intelligent tip that based on user interest model, relevant search and other auxiliary search function were developed. Finally, two sets of experiments were carried to test the recall, precision and response time and other indicatorsoe the engine. Through which we make sure that the ontology-based semantic similarity search technology of geographic information is reasonable.