Researchers at the University of Pennsylvania discovered that Twitter keyword data can be analyzed to show geographic patterns similar to heart disease rates mapped by the Centers for Disease Control and Prevention.
Researchers took a sample of 826 million tweets from June 2009 to March 2010 and analyzed negative and positive words to map a tone for each community. They found that analyzing the sentiments trumpeted over Twitter was a technique more predictive and inexpensive than traditional methods of identifying regional heart disease. Twitter has grown in popularity since 2010, which researchers said would make the technique even more accurate today.
