Automatic mapping method of place name categories considering context information

Different platforms use different toponymic classification systems and the existing toponymic classification systems containing global toponymic data are all non Chinese, which poses a challenge for cross platform data sharing and integration. In order to effectively use these toponymic data resources, it is first necessary to achieve category mapping in different toponymic classification systems. This paper constructs a generic Chinese and English database on the premise of mapping non Chinese toponymic classification systems to Chinese toponymic classification systems. It proposes an automatic category mapping method that takes into account the contextual information of toponymic categories. It uses the toponymic category context to obtain the category and generic matrix. The mapping mode is one to many and many to one. It calculates the similarity basis between the original category and the target category. According to the frequency of generic names, it takes the higher similarity value as the category mapping result, so as to achieve data standardization and interoperability. Operability. Through experimental verification, the classification performance of our method on large-scale datasets is significant, which can achieve standardization of Chinese platform data in place name classification system and improve the compatibility and usability of data across multiple systems.