How tweets can predict happiness: Penn’s World Well-Being Project

What you tweet says a lot about how happy you are, according to a group of researchers at Penn.

The World Well-Being Project, a Penn research group that analyzes language used in social media to better understand life satisfaction, looked at millions of geocoded tweets from 2009 to 2010 for this study, which showed that certain words are accurate markers of high or low life satisfaction. Previously, scientists had ideas about how certain words could be linked to happiness or unhappiness, but there was no way to prove it, said the project’s lead research scientist Andy Schwartz.

In this study, the team identified words from tweets that it thought could signify high or low levels of happiness and then compared its findings to a Gallup poll on life satisfaction. The Twitter word map and the Gallup map matched, the researchers found, suggesting that certain words can indeed predict happiness or unhappiness.

Why use Twitter? It’s “the best source of publicly available data,” Schwartz said, and it’s possible to geocode.

Read more about the study in this blog post by Schwartz.

And yep, in that map, above, Philly is marked as having a very low life satisfaction (as per the Gallup poll and the World Well-Being Project’s work).