Communications of the ACM
Could Social Media Be sed to Detect Disease Outbreaks?
Credit: Medindia
University of Bristol researchers are studying whether social media can be used to track an event or phenomenon, such as flu outbreaks and rainfall rates. Bristol professors Nello Cristianini and Vaseleios Lampos geo-tagged user posts on Twitter as the input data to investigate flu-like illness rates and rainfall levels.
The research is based on an early study involving Flu Detector, a tool that uses Twitter content to map current flu rates in several United Kingdom (UK) regions. "We were able to turn geo-tagged user posts on the microblogging service of Twitter to topic-specific geo-located signals by selecting textual features that showed the content and understanding of the text," Cristianini says.
The researchers used machine-learning algorithms to automatically determine which keywords in the database of tweets were associated with increased flu levels. Over several months the researchers gathered more than 50 million geo-located tweets, comparing them to data from the UK's National Health Service on flu incidence by region.
From University of Bristol
View Full Article
Abstracts Copyright © 2011 Information Inc., Bethesda, Maryland, USA
No entries found