Abstract:The construction of emergency warning information release system is an important means to prevent disaster risk. According to the characteristics of emergency early warning information and early warning information data in Hainan Province, carry out system demand analysis, propose the system structure of emergency early warning information release system, use big data technologies such as data mining and data simulation to assist emergency decision-making, analyze the development trend, and develop emergency plans, evaluate the early warning effect, and dynamically adjust emergency strategies. The overall architecture of the system is divided into four levels: business application, platform service, data support and infrastructure, and a big data early warning information release model is designed. The system functional framework is divided into three subsystems: real-time collection, early warning information release and early warning information release channel. Through the comparative test of the collection, data processing and release performance of the early warning information of the original system and the new system, the response capacity of the Hainan emergency early warning information release system based on the big data platform has increased by 1.8 to 2.5 times.