利用地理知识图谱的COVID-19疫情态势交互式可视分析

Interactive Visual Analysis of COVID-19 Epidemic Situation Using Geographic Knowledge Graph

新型冠状病毒肺炎(coronavirus disease 2019,COVID-19)疫情的暴发产生了海量与时空信息相关的数据,当前的地理时空疫情分析难以关联人物关系、事件等数据,由此提出了一种利用地理知识图谱结合交互式可视分析COVID-19疫情态势的方法。首先定义了病患实体和关系类型,提出了事件语义模型和事件关系分类,并根据不同的数据分类设计了知识抽取和知识表示方法,构建了病患时空信息知识图谱;然后分别从宏观和微观层次把控疫情态势的任务出发,设计了知识图谱可视化视图和交互分析方法;最后构建实验分析系统,利用COVID-19确诊患者数据,通过地图分布可视化、图谱可视化和轨迹可视化等多视图协同交互对COVID-19疫情态势进行分析实验。实验结果表明,该方法能够从实时疫情态势监控、病患关联关系、高危人群防控和地区防控态势等方面为疫情态势分析提供一种新的思路与方法。

After the outbreak of coronavirus disease 2019(COVID-19),there are large amounts of data related to spatiotemporal information.It is difficult for the epidemic analysis with geographic spatiotemporal model to take the data of character relationship and events into consideration.Therefore,a method of using geographic knowledge graph and interactive visualization to analyze the epidemic situation of COVID-19 is proposed.First,the entity and relationship types of patients are defined,the event semantic model and event relationship classification are proposed,and according to different types of data,knowledge extraction and knowledge representation methods are designed,and the knowledge graph of patients’spatiotemporal information is constructed.Then,from the macro and micro level of the task of epidemic situation control,an analysis frame combining semantic network and visual analysis model is proposed.Finally,an experimental analysis system is built,which uses the data of confirmed patients from COVID-19 to carry out the experiment on the situation analysis of COVID-19 through multi-view collaborative interaction analysis such as map distribution visualization,knowledge graph visualization and trajectory visualization.The experiment has proved that it can provide a new way to analyze the epidemic situation from the aspects of realtime situation monitoring,patient relationship analysis,high-risk population analysis and regional control situation analysis.