An Approach Of Semantic-based Integration On Geographic Information

Even for the same object, data types from different Geographic Information Systems (GIS) will vary from each other. To offer users more comprehensive and integral geographical information, it is necessary to integrate different data sources. With the development of internet, there has been increasing geographical information data in Web and other data sources. Generally speaking, the data types from divergent data sources are heteroid, and there exist lots of problems in the data integration of various data sources. The most crucial is how to classify different data sources according to their characteristics and how to map the geographical information between differentiated data sources. With the adoption of semantic-based characteristic classification and definition methods, the paper presented a geographical information acquisition model with semantics as the core. In this way, it offered a solution to the semantic heterogeneity of distributed geographical information sources. Besides, the model was able to integrate the geographical information from the collected semantic data through integrated approaches.In view of the heterogeneity and the difficulty in ensuring the accuracy during the data integration of existing multi-source geographic information, the paper began with analysis of current geographic information download methods in foreign internet-related electronic map websites. In this way, relevant geographic data was obtained, and data analysis was made of the geographic information, with original data necessary for researches becoming available. Secondly, the paper studied the data characteristics classification systems for different geographic information, and carried out semantic mapping and elimination of heterogeneity on characteristics between data sources. Thirdly, with the adoption of a kind of multi-feature fusing data matching method, the paper measured the similarity of different data, so as to link relevant and similar data and achieve the integration of multi-source geographic information. The paper ended up with experimental analysis of the integrated data with relevant assessment systems. According to the experiment results, the similarity calculation methods presented in the paper were highly accurate. This has laid a solid foundation for the integration of geographic information.