Abstract: Existing emergency intelligence support has some deficiencies, such as static knowledge, fuzzy reasoning, and spatial homogeneity. The paper proposes a path of emergency intelligence support that integrates eventic graph and network opinion analysis, aiming to improve the situational awareness and predictive response capability of emergencies in complex situations, and support emergency intelligence paradigm innovation in the era of digital intelligence empowerment. The paper constructs a framework for emergency intelligence support from the integrated perspective of reasoning and knowledge fusion, business and network fusion, and carries out empirical research using hazardous chemical accidents as an illustration. The work includes building a database of relevant emergencies, portraying the development logic of emergencies using an eventic graph, and describing the impact of emergencies in cyberspace using online opinion analysis techniques. This provides intelligence support for emergency management and decision making for new emergencies. The intelligence support path proposed in this paper can take into account the characteristics of multi-source spatial data as well as the integration of event logic and knowledge, allowing for a more systematic "past-future" intelligence prediction of emergencies. The empirical study and validation of hazardous chemical accidents also show good knowledge presentation and intelligence reasoning.

Key words: Emergencies, Emergency intelligence, Eventic graph, Network public opinion, Causal relations