Monday, April 15, the day of the
Scientists at the University of Vermont and MITRE Corp. have been crunching millions of messages from Twitter in an effort to quantify the public mood. Their results went public Tuesday at hedonometer.org
In February, the gang at the University of Vermont and MITRE made headlines when it declared the happiest and saddest cities in the U.S., based on geo-tagged tweets from cellphones: Napa, Calif., and Beaumont, Texas. You guess which was which.
"Reporters, policymakers, academics — anyone — can come to the site and see population-level responses to major events," said Chris Danforth, a Vermont mathematician who co-led the creation of the site with fellow mathematician Peter Dodds.
Makers of the app say they soon will branch out from tweets in English to data in 12 languages from various streams:
Shortly after two
Among the negative words that trended that day were "explosion," "victims" and "kill." Mathematical models recognized that they stuck out from the normal "ambient chatter."
The hedonometer ticker starts with measurements of the "psychological valence" of some 10,000 words, using
The results were used to rate about 50 million daily tweets. The method works because of the scale – erratic uses and contexts tend to wash out, and accurate trends emerge.
"It's the relative context that is so important, which is why the sudden drop from the Boston Marathon bombings jumps out at you," said Brian Tivnan, a researcher from MITRE, a nonprofit that operates federal research and development centers. "The hedonometer shows the pulse of a society."