Penn’s “Twitter Predicts Heart Disease” Team

The World Well-Being Project (WWBP) is a multidisciplinary research group at the University of Pennsylvania. A collaboration between computer scientists and psychologists, WWBP is pioneering big data techniques for measuring physical and psychological health and well-being based on language in social media.

In a study released in January of 2015, led by graduate student Johannes C. Eichstaedt, the Penn team found that Twitter can predict community mortality rates from heart-disease better than 10 common demographic, socioeconomic, and health risk factors, including smoking and hypertension.

Specifically, language patterns reflecting negative social relationships, disengagement, and negative emotions—especially anger—emerged as risk factors; positive emotions and psychological engagement emerged as protective factors. WWBP’s work suggests that Twitter is not only a meaningful indicator of community well-being, but that big data language analysis can complement, and in part replace, traditional survey methods.

Read more about the project at wwbp.org.