Twitter is proving to be a viable way to track seasonal flu outbreaks, though researchers recommend further study regarding using tweets to monitor influenza outbreaks.
A study by nearly a dozen San Diego State University researchers, published in the Journal of Medical Internet Research, reveals the social network is becoming a more reliable and accurate supplementary surveillance tool when it comes to identifying flu outbreaks. "With the popularity of social media growing, the Internet is a source for syndromic surveillance due to the availability of large amounts of data," according to the study abstract, which notes that collecting data on flu activity can be slow, and typically delayed by up to two weeks, when using traditional data collection methods.
The study's three-pronged goal was improving the correlation of tweets using filtering and a machine-learning classifiers, observing tweets regarding emergency department influenza-like illness and tweets indicating laboratory-confirmed flu cases in San Diego. Researchers collected tweets with the keyword "flu" in 11 U.S. cities, totaling 159,802 tweets. Article