特定关键词百度指数与新型冠状病毒肺炎流行相关性分析

Correlation analysis between Baidu index of specific keywords and the prevalence of COVID-19

目的分析特定关键词的百度指数与新型冠状病毒肺炎流行的相关性,探讨大数据时代中新发传染病监测的新思路。方法以范围选词法为主,直接选词法为辅,结合国家新冠肺炎诊疗及防控方案,并运用百度指数的拓展功能确定关键词库;选取2020年1月19日至2月29日的数据,分别对每个关键词的百度指数与新冠肺炎发病数进行Spearman相关分析。结果共有25个关键词进入词库,"口罩"、"N95口罩"、"外科口罩和医用口罩的区别"等15个关键词的百度指数与新冠肺炎发病数的相关关系有统计学意义(P<0.01),其中"发热"、"乏力"、"咳嗽"等11个关键词的百度指数序列向前移动若干单位时间后,与新冠肺炎发病数的相关系数达到最大值,相关系数为0.71~0.92。结论"发热"、"乏力"、"咳嗽"等11个关键词作为先行关键词可提前对新冠肺炎的疫情做出反馈,具有一定预测预警的价值。

Objective To analyze the correlation between Baidu index of specific keywords and the prevalence of COVID-19,and to discuss the new ideas for monitoring emerging infectious diseases in the era of big data.Methods Based on the scope of lexical selection,supplemented by direct lexical selection,combined with the national COVID-19 diagnosis and treatment and prevention and control program,and the use of Baidu index expansion function to determine the keyword library;the data from January 19 to February 29,2020 were selected,and the Spearman correlation analysis was performed on the Baidu index of each keyword and the number of cases in COVID-19.Results A total of 25 keywords were selected into the words library.The correlations of the novel coronavirus pneumonia incidence and Baidu index of"masks","N95 mask","surgical mask and medical mask"and other 15 key words,were statistically significant(P<0.05).Among them,Baidu index of the 11 key words,"fever","fatigue","cough",moved forward a certain number units,reaching the maximum correlation coefficient with the number of novel coronavirus pneumonia cases.The correlation coefficients was 0.71-0.92.Conclusion The 11 key words including"fever","asthenia"and"cough"as the leading key words could provide feedback on the epidemic situation in COVID-19 in advance,and had certain prediction and early warning value.