基于图数据库的流调数据轨迹追踪技术的建立

职业与健康 ›› 2023, Vol. 39 ›› Issue (20): 2819-2825.

基于图数据库的流调数据轨迹追踪技术的建立

何海艳¹, 王钰铭², 王永馨¹, 张国平¹, 赵莹¹, 吴伟慎¹

1.天津市疾病预防控制中心传染病预防控制所,天津 300011;
2.天津医科大学公共卫生学院,天津 300007

收稿日期:2023-02-19 修回日期:2023-03-14 发布日期:2026-03-26
通信作者: 吴伟慎,主任医师,E-mail：wuweishen@live.cn
作者简介:何海艳,女,副主任医师,主要从事传染病预防与控制工作。
基金资助:
天津市卫生健康科技项目（ZC20019）; 天津市医学重点学科（专科）建设项目（TJYXZDXK-050A）; 天津市公共卫生科技重大专项（21ZXGWSY00010）

Establish of trace tracking technology of streaming data based on graph database

HE Haiyan¹, WANG Yuming², WANG Yongxin¹, ZHANG Guoping¹, ZHAO Ying¹, WU Weishen¹

1. Infectious Disease Prevention Control Room,Tianjin Centers for Disease Control and Prevention,Tianjin 300011,China;
2. School of Public Health,Tianjin Medical University,Tianjin 300007,China

Received:2023-02-19 Revised:2023-03-14 Published:2026-03-26
Contact: WU Weishen,Chief physician,E-mail：wuweishen@live.cn

摘要/Abstract

摘要： 目的利用重大突发公共卫生事件流调数据中的感染者活动轨迹信息,实现感染者活动轨迹的可视化。方法基于Neo4j图数据库的深度挖掘流调数据的方法,利用2020年初天津市某远郊区重大突发公共卫生事件中感染者流调数据,使用NLP领域的文本分类和命名实体识别等相关深度学习技术,将流调数据拆解为感染者、感染类型、地点、地点类型、活动事件及活动事件类型6类实体及其实体关系,实现对感染者活动轨迹的追踪,辅助划定风险区域。结果本研究选择任意时间段内60例感染者的聚集度,最终形成节点和关系数据7 390条数据,其中节点1 643条,关系5 747条,得出的结果与实际聚集区域基本一致。经过2022年天津市某近郊区疫情数据进一步校验,根据流调数据中感染者活动轨迹和活动范围,可生成感染者活动热力图与官方封控的区域基本吻合。结论感染者的热力图分布可以评估疫情可能波及的范围,辅助划定风险区域,有利于进一步精准的流行病学调查和政府的精准管控。

关键词: 重大突发公共卫生事件, Neo4j图数据库, 轨迹追踪, 可视化分析

Abstract: Objective To realize the visualization of the activity trajectory of infected persons by the activity trajectory information of infected persons in the major public health emergencies surveillance data. Methods Based on the method of deep mining of Neo4j graph database,the traffic survey data of major public health emergencies infected persons in outer suburb of Tianjin in early 2020 were used to implement data mining. Using text categorization and named entity recognition in the NLP deep learning technology,the traffic survey data were decomposed into six types of entities and their entity relationships,including infection,infection type,location,location type,activities and activities type,in order to track the movement track of the infected person and assist in delineating the risk area. Results In this study,the aggregation degree of 60 infected persons in any time period was selected,and finally 7 390 pieces of node and relationship data were formed（including 1 643 nodes and 5 747 relationships）,and the results were basically consistent with the actual clustering area. Furthermore,the data of epidemic situation in near suburb of Tianjin in 2022 were further verified. According to the data of epidemic survey,the activity locus and activity range of infected persons could be generated,and the activity heatmap of infected persons was basically consistent with the area sealed off by the government. Conclusion The distribution of the heat map of infected persons can be used to assess the possible spread of the epidemic situation and help to delineate the risk areas,which is beneficial to the further accurate epidemiological investigation and the precise control of the government.

Key words: Major public health emergencies, Neo4j diagram database, Trajectory tracking, Visual analysis

中图分类号:

R181.2

何海艳, 王钰铭, 王永馨, 张国平, 赵莹, 吴伟慎. 基于图数据库的流调数据轨迹追踪技术的建立. [J]职业与健康, 2023, 39(20): 2819-2825.

HE Haiyan, WANG Yuming, WANG Yongxin, ZHANG Guoping, ZHAO Ying, WU Weishen. Establish of trace tracking technology of streaming data based on graph database. [J]OCCUPATION AND HEALTH, 2023, 39(20): 2819-2825.

参考文献

[1] 陈莹,郭怡博,郭然,等.基于文献计量学的新型冠状病毒肺炎(covid-19)研究可视化分析[J].中国中药杂志,2020,45(10):2239-2248.
[2] ZHENG W,YAN L,GOU C,et al.Pay attention to doctor-patient dialogues:Multi-modal knowledge graph attention image-text embedding for covid-19 diagnosis[J].Inf Fusion,2021,75:168-185.
[3] CHEN T,ZHANG Y,QIAN X,et al.A knowledge graph-based method for epidemic contact tracing in public transportation[J].Transp Res Part C Emerg Technol,2022,137:103587.
[4] PESTRYAKOVA S,VOLLMERS D,SHERIF M A,et al.Covidpubgraph:A fair knowledge graph of covid-19 publications[J].Sci Data,2022, 9(1):389.
[5] 中华人民共和国国家卫生健康委员会.新型冠状病毒肺炎诊疗方案(试行第六版)[Z].北京,2020:1-13.
[6] 中华人民共和国国家卫生健康委员会.新型冠状病毒肺炎防控方案(第八版)[Z].北京,2020:1-12.
[7] QUIROZ J C,CHARD T,SA Z,et al.Extract,transform,load framework for the conversion of health databases to OMOP[J].PLoS One,2022, 17(4):e0266911
[8] GOKHALE K M,CHANDAN J S,TOULIS K,et al.Data extraction for epidemiological research(dexter):A novel tool for automated clinical epidemiology studies[J].Eur J epidemiology,2021,36(2):165-178.
[9] 庞东旭,王广舜,门学博,等.天津宝坻区58例新型冠状病毒肺炎聚集发病流行病学调查[J].中国公共卫生,2020,36(3):289-292.
[10] OTTER D W,MEDINA J R,KALITA J K.A survey of the usages of deep learning for natural language processing[J].IEEE transactions on neural networks and learning systems,2021,32(2):604-624.
[11] 李冬梅,罗斯斯,张小平,等.命名实体识别方法研究综述[J].计算机科学与探索,2022,16(9):1954-1968.
[12] WU B,WEI B,LIU J,et al.Faceted text segmentation via multitask learning[J].IEEE Trans Neural Netw Learn Syst,2021,32(9):3846-3857.
[13] 余同瑞,金冉,韩晓臻,等.自然语言处理预训练模型的研究综述[J].计算机工程与应用,2020,56(23):12-22.
[14] 施国良,谢泽宇,杨小莉.高校图书馆复杂网络构建与智慧化应用探索[J].图书情报工作,2019,63(23):106-112.
[15] 李佳,刘海砚,刘俊楠,等.基于知识图谱的新冠肺炎病例传播关系可视分析[J].信息工程大学学报,2021,22(5):606-612.
[16] WU J.Construct a knowledge graph for china coronavirus(covid-19) patient information tracking[J].Risk Manag Healthc Policy,2021,14:4321-4337.
[17] 蒋秉川,游雄,李科,等.利用地理知识图谱的covid-19疫情态势交互式可视分析[J].武汉大学学报(信息科学版),2020,45(6):836-845.
[18] 吴伟慎,李永刚,魏兆飞,等.天津市某百货大楼新型冠状病毒肺炎聚集性疫情调查分析[J].中华流行病学杂志,2020,41(4):489-493.
[19] 潘静静,于铭洋,李亚飞,等.河南省首起奥密克戎新冠肺炎疫情流行特征分析[J].现代预防医学,2022,49(14):2656-2660.
[20] 张国平,苏承,杨杰,等.新型冠状病毒omicron变异株引发天津市家庭聚集性疫情的传播特征及影响因素[J].中华流行病学杂志,2022, 43(9):1370-1375.
[21] 马艺文,潘琛玲,孟成.武汉市居住空间特征与covid-19疫情分布关系研究[J].地理空间信息,2020,18(11):43-47.
[22] 王成文,熊励.基于知识图谱的突发公共卫生事件辅助诊疗研究[J].情报科学:2022,9(1):1-12.
[23] 张雷雨,刘杰.Covid-19地理知识图谱构建分析——以江苏省为例[J].连云港职业技术学院学报,2022,35(1):6-11.
[24] JIANG B,YOU X,LI K,et al.Interactive analysis of epidemic situations based on a spatiotemporal information knowledge graph of covid-19[J].IEEE Access,2022,10:46782-46795.