Objectives   The knowledge graph is an important tool for revealing entities and the relationships between them. The role of knowledge graph in integrated disaster reduction is becoming increasingly prominent. We summarized the knowledge graph construction method and application for integrated disaster reduction.

  Methods   Firstly, the concepts of knowledge graph and the application of knowledge graph in disaster reduction is introduced. The knowledge graph can realize the rapid aggregation of multi-source heterogeneous data in the integrated and comprehensive disaster reduction, and organize the data of the relevant departments in an orderly manner. The knowledge graph can establish relationships between social fields related to emergency rescue, and reveal cross-network relationships between different fields, different social entities, and entities and data resources and disaster events. Knowledge graphs can more efficiently extract and utilize time-sensitive and information-intensive Internet and social media data.Secondly, the knowledge graph construction process and key technologies for integrated disaster reduction are summarized. Specifically, the knowledge graph construction process includes process multi-source heterogeneous data, extracting entities and relationships from the data according to the application scenario,fusing various types of know- ledge, and finally modeling knowledge graph and store it in the knowledge base. The key technologies mainly include knowledge extraction,information fusion,knowledge graph building,and knowledge storage.

  Results   The knowledge graph has established the connection between the user and the required information, and personalized information can be pushed to three types of users based on the knowledge graph. The main users of the system include three categories, namely emergency management users, public users and emergency rescue users.

  Conclusions   The knowledge graph has the advantage of gathering multi-source heterogeneous data, displaying rich disaster related information, pushing personalized information. The degree of automation of knowledge graph construction in the field of emergency disaster reduction is insufficient. Massive structured and unstructured data brings challenges to the storage and rapid construction of knowledge graphs. In the field of disaster reduction and emergency response, how to effectively and uniformly manage various types of earthquake information in practical applications, improve the prediction accuracy of disaster development trends, and discover the temporal and spatial patterns, evolution laws, activity patterns and internal mechanisms of disasters still need to be further expanded and deepened.