基于在线健康社区用户信息行为的疫情风险识别模型研究--以“COVID-19”为例

Research on Epidemic Risk Identification Model Based on Users Information Behavior in Online Health Community:Case Study of“COVID-19”

[目的/意义]为利用在线健康社区问诊数据进行疫情风险感知,提出一种基于在线健康社区用户信息行为的疫情风险识别模型,拓宽疫情风险的识别渠道和感知能力。[方法/过程]对《中华人民共和国传染病防治法(2013修正版)》规定的甲、乙、丙类37种传染病进行科学知识图谱分析、术语提取及补充后,得到不重合症状142种,利用GRA找出每种传染病关联度排名较高的症状作为需要识别的风险要素以建立症状风险指标;对在线健康社区用户生成内容进行分析来构建其余指标,与症状风险一同构成本文疫情风险指标体系,并通过基于预测的ARIMA构建风险识别模型。[结果/结论]通过对“COVID-19”爆发初期的在线健康社区数据仿真结果表明,提出的基于GRA-ARIMA多维时间序列组合模型具备较优的疫情风险识别能力,模型能够基本拟合“COVID-19”疫情初期发展曲线,可为疫情二次爆发的提前识别及我国疫情风险感知神经末梢的建立健全提供参考。

[Purpose/significance]In order to use online health community consultation data for epidemic risk perception,an epidemic risk identification model based on online health community user information behavior is proposed to broaden the identification channel and perception ability of epidemic risk.[Method/process]After carrying out analysis on the scientific knowledge map,terminology extracting and supplementing the terms of 37 infectious diseases of class A,B and C specified in the Law of the People’s Republic of China on the Prevention and Treatment of Infectious Diseases(2013 Revised Edition),142 inconsistent symptoms are obtained.The symptoms with high correlation degree of each infectious disease are found by GRA as the risk elements to be identified,so as to establish the symptom risk index;based on the analysis on the content generated by the users from online health community,the other indicators are built,the epidemic risk index system consist with the symptom risk and the other indicators,and the risk identification model ARIMA is constructed based on prediction.[Result/conclusion]The simulation results of online health community data at the initial stage of“COVID-19”outbreak show that the multi-dimensional time series combination model based on GRA-ARIMA proposed in this paper has better identification ability in epidemic risk,can basically fit the initial development curve of“COVID-19”epidemic,therefore,it can provide a reference for the early identification of the second outbreak and the establishment and improvement of epidemic risk sensing nerve endings in China.