Using Social Media to Model H1N1

By Clive Riddle, September 10, 2009

In the September 2009 issue of Predictive Modeling News, Russell A. Jackson reports on the use  of Twitter, Facebook, blogs, search engines and more in modeling the spread of the H1N1 influenza virus, in his article “Social Media, Traditional Data Sources Fuel Swine Flu Models” Both social media updates and comments, and keywords from online searches, can provide data useful for such analytics.

Russell writes “One example of the cutting-edge marriage of predictive modeling and social media is the Social Web Information Monitoring for Health – or ‘SWIM for Health’ -- project operating out of the University of Iowa. Researchers there have embarked on a major study that tracks public perception of the swine flu outbreak and other infectious diseases, utilizing technology from OneRiot that indexes the social web in real time. The project, its participants say, ‘has the potential to enhance disease tracking and forecasting by harnessing the power of the social web.’ By monitoring updates from Twitter and Facebook, recent blog posts, current popular search queries and other web usage activity, public health officials can potentially locate an influenza outbreak or simply indicate an elevated perception of disease risk. Such information might help public health authorities better address public concerns. The first step in the research includes an interactive, real-time map of the United States that monitors swine-flu-related Twitter updates. It’s available at http://compepi.cs.uiowa.edu/swim/.

Who is OneRiot you ask? Jackson tells us “launched in November 2008, OneRiot “finds news, stories and videos people are talking about right now across the social web.” Unlike any other search engine, it ranks a web page’s relevance based on its current popularity with real people. OneRiot is a privately held company headquartered in Boulder, CO, with offices in San Francisco.”

Another initiative Russell cites is a project in which “researchers with the National Institutes of Health’s Models of Infectious Disease Agent Study, or ‘MIDAS,’ recently posted these questions on Facebook: [1} “When did you first learn about the swine flu outbreak?”  2} “Have you searched the internet for additional information on the swine flu outbreak?” {3} “If a vaccine for swine flu became available, would you want to be vaccinated?” …. The researchers will use the Facebook responses to build a dynamic model that simulates how changes in decision-making influence patterns of disease spread. The model will help them and others identify the strategies that improve adherence to interventions and reduce the spread of disease. “

So perhaps not only can you reduce your chance of contract with H1N1 by socializing online instead of in person, but you might help measure and combat its spread at thie same.