Construction and Emergency Response Application of Knowledge Graph

Emergency response preparedness is the core element that enhances the effectiveness of emergency management, and it possesses comprehensiveness, systematicity, authority, and practicality. This paper aims to improve its data support and preventive disposal capability through artificial intelligence, especially using Neo4j and Python, multi-feature fusion, Tokenizer word segmentation, direct mapping method, manually defined rules, long and short term memory network(LSTM), and content analysis to construct a knowledge graph. After knowledge extraction, the raw data was structured and 19 kinds of entities with 15 kinds of relationships were extracted. Through knowledge fusion and knowledge storage, the knowledge graph of emergency plans is successfully constructed. This system not only helps managers to optimize inventory management and demand forecasting, but also improves the efficiency of the supply chain and reveals the correlation between materials.