Abstract: Government information sharing service is the great significance for building a service-oriented government and modernizing the country's governance system and governance capabilities. At present, government information resources in various industries and departments are in various forms and rich in content. And most of them are related to geographic information, Which is mainly stored through address fields. However, duing to different collection calibers, collection time, collection standards, etc., the description of address information is inconsistent. And it is difficult for the government information of different departments to be aggregated to a unified geographic location. Or after the information is aggregated, data mining and analysis of multiple data sources cannot be carried out duing to the single type of aggregation. Therefore, in this paper, the construction of the smart Xinyu space-time information cloud platform project is taken as an example, and a method based on standard geographical name and address matching is proposed for integrating government information resources, which improves the integration efficiency and shared development level of government information. Moreover, after integration with geographic entities, government information can be "planned with graphs, analyzed with graphs, and written with graphs", which can promote cross-level, cross-regional, cross-industry, and cross-departmental sharing of government information resources.

Key words: government information resources, information resource integration, big data, standard geographical name and address, geographic entity, smart Xinyu