Disease spread is a complex phenomenon requiring an interdisciplinary approach. Covid-19 exhibited a global spatial spread in a very short time frame resulting in a global pandemic. Data of web search effort in Greece on Covid-19 as a topic for one year on a weekly temporal scale were analyzed using governmental intervention measures such a s school closures, movement restrictions, national and international travelling restrictions, stay at home requirements, mask requirements, financial support measures, and epidemiological variables such as new cases and new deaths as potential explanatory covariates. The relationship between web search effort on Covid-19 and the 16 in total explanatory covariates was analyzed with machine learning. Web search in time was compared with the corresponding epidemiological situation, expressed by the Rt at the same week. Results indicated that the trained model exhibited a fit of R2 = 91% between the actual and predicted web search effort. The top five variables for predicting web search effort were new deaths, the opening of international borders to non-Greek nationals, new cases, testing policy, and restrictions in internal movements. Web search peaked during the same weeks that the Rt was peaking although new deaths or new cases were not peaking during those dates, and Rt rarely is reported in public media. As both web search effort and Rt peaked during 1-15 August 2020, the peak of the tourist season, the implications of this are discussed.
翻译:Covid-19在极短的时间内展示了全球空间分布,导致出现全球流行病; 每周按时间规模对希腊Covid-19作为Covid-19专题的网络搜索数据进行了为期一年的每周时间范围分析; 利用政府干预措施,如学校关闭、行动限制、国家和国际旅行限制、留在家中的要求、面具要求、财政支助措施、流行病变量,如新病例和新死亡等潜在的解释性共变体; 以机器学习对Covid-19和16个解释性共变体的网络搜索努力之间的关系进行了分析; 及时将Covid-19的网络搜索工作与Rt在同一周表达的相应流行病状况进行了比较; 结果表明,经过培训的模式在实际和预测的网络搜索工作之间显示出91%的适合度; 预测网络搜索工作的最大五个变量是新的死亡、向非希腊国民开放国际边界、新案例、测试政策和内部流动的限制; 在Rt高峰期的几周内,网络搜索工作达到高峰,尽管Rt是新的死亡或新案例,但在同一星期内,在2020年8月1日的旅游高峰期间,统计高峰期的搜索工作没有达到高峰期。