Research on the Construction of Knowledge Graph for Emergency Management of Social Security Events

Constructing a knowledge graph for social security event emergency management promotes the development of social security event emergency management towards an intelligent model that relies on a vast amount of knowledge and data. Relevant content is collected from multiple information sources, and the UIE model, fine-tuned with a small amount of annotated data, is utilized to achieve joint entity-relation extraction. The results are stored in a Neo4j graph database, laying the foundation for application and analysis. The research concludes that utilizing LLM-KG fine-tuning provides a feasible means for constructing large-scale knowledge graphs in vertical domains with low resources, The knowledge graph obtained from this research can satisfy application scenarios in many situations.