Scientometric Trends for Coronaviruses and Other Emerging Viral Infections

Dima Kagan; Jacob Moran-Gilad; Michael Fire

Data about infectious diseases can also come from web-and social-based sources. For instance, in 2009, Ginsberg et al. [12] used Google search queries to monitor the spread of the influenza epidemics. They used the fact that many people search online before going to doctors, and they found that during a pandemic, the volume of searches differs from normal. They then created a mathematical model to forecast the spread of flu. This research was later converted into a tool called Google Flu Trends, and at its peak, Google Flu Trends was deployed in 29 countries worldwide. However, not everything worked well for Google Flu Trends; in 2009, it underestimated the flu volume, and in 2013, it predicted more than double the number of cases than the true volume [13] . As a result of such discrepancies, Google shut down the Google Flu Trends website in 2015 and transferred its data to academic researchers [14] . Also in 2009, Carneiro and Mylonakis [15] used large amounts of data to predict flu outbreaks a week earlier than prevention surveillance systems. Similar to Ginsberg et al. [12] , they used Google Trends as a data source, but instead of only monitoring seasonal flu, they monitored avian influenza and West Nile virus. Since there was a constant growth in the number of Google searches, they normalized the data to get meaningful results. They found a correlation between web searches about influenza and the CDC data.