摘要
为了预判公共卫生事件演化态势,提出基于事理图谱的公共卫生事件次生衍生事件预警模型。利用2层本体实现结构化情景表示,综合利用预训练模型与深度神经网络实现事件抽取,运用模式匹配的方法抽取事件逻辑关系,融合词嵌入模型与聚类算法进行事件泛化,进而实现事理图谱构建,运用相似度计算与逻辑预测实现次生衍生事件预警,并以“甲型流感”和“肺炎支原体肺炎”事件为例进行实证研究。研究结果表明:“甲型流感”事理图谱能够清晰展示该事件与相关事件的逻辑关联,基于该事理图谱能够实现“肺炎支原体肺炎”事件的次生衍生事件预警。研究结果可为公共卫生事件次生衍生事件预防工作提供参考。
To predict the evolution situation of public health emergencies,an early-warning model for the secondary and derivative events of public health emergencies based on the event evolutionary graph was proposed.The two-layer ontology was utilized to achieve the structured scenario representation,the pre-training model and deep neural network were comprehensively applied to realize the event extraction,then the pattern matching approach was employed to extract the logical relationship of events,and the word embedding models and clustering algorithms were integrated to achieve the event generalization,thereby realizing the construction of event evolutionary graph,the similarity calculation and logical prediction were performed to achieve the early-warning for the secondary and derivative events.Moreover,the empirical research was conducted by employing“Influenza A virus”and“M.Pneumonia”events as examples.The results show that the“Influenza A virus”event evolutionary graph can clearly display the logical correlation between this event and the related events,and based on this graph,the secondary and derivative event early-warning for the“M.Pneumonia”event can be achieved.The research results can provide reference for predicting the secondary and derivative events of public health emergencies.
作者
李诗轩
LI Shixuan(School of Safety Science and Emergency Management,Wuhan University of Technology,Wuhan Hubei 430070,China)
出处
《中国安全生产科学技术》
CAS
CSCD
北大核心
2024年第3期12-19,共8页
Journal of Safety Science and Technology
基金
国家自然科学基金青年科学基金项目(72204194)
湖北创新发展研究院开放课题(CX2023-1-2)
关键词
公共卫生事件
事理图谱
次生衍生事件
相似度计算
可解释性预警
public health emergencies
event evolutionary graph
secondary and derivative events
similarity calculation
interpretable early-warning
