谷歌流感趨勢的成功與失誤

Opportunities and challenges coexist in the big data era, how to handle big data based on the traditional methods is thought-provoking, blind pursuit of big data is not necessarily correct. This paper discusses Google Flu Trends (GFT) as an example to introduce the main techniques and achievements in disease surveillance with big data, and provides some improvements with sophisticated statistical tools. The success of GFT depends on the application of correlation, while its failure is derived from model errors, conflict between causation and correlation and other issues. Analysis and enlightenment of GFT are of theoretical and practical significance for government to provide big data solutions in the future.