Where life is tweet: Study ranks #happy places
http://www.usatoday.com/story/news/nation/2013/02/20/happiest-cities-geotags-twitter/1933587
#:-) vs #:-(
Où trouve-t-on le bonheur et la tristesse (aux É.-U.)
Analyse du vocabulaire de 10 millions de tweets.
The study, from mathematicians at the University of Vermont, is the latest to tap Twitter’s 140-character, real-time posts to see how we’re feeling. Previous studies found that Twitter users are happiest on Sunday mornings, saddest on Thursday evenings and prone to perk up on Christmas and other holidays.
The new study used a list of 10,000 words rated on a 10-point scale as happy, sad or neutral to score tweets from 2011 that carried geographic tags. The researchers threw out neutral words (such as ‘the,’ “of” and ‘and’) and looked at how often the happy and sad words showed up in different cities and states, lead researcher Lewis Mitchell says.
Words such as ‘hate,’ “terrorist,” ‘earthquake’ and ‘greed’ were high on the sad list, he says. Happy words included ‘happy,’ “reunion,” ‘lol’ (laughing out loud) and nature terms, which helps explain how tweets from Maine, which mentioned lots of ‘forests’ and ‘rivers,’ came in only second to those from Hawaii on the happy list.
Other relatively happy states included Nevada, Utah and Vermont. After Louisiana, the least happy states were Mississippi, Maryland, Delaware and Georgia.
http://www.uvm.edu/storylab/share/papers/mitchell2013a/AppendixB_files/states_2011.jpg
Figure 1: Choropleth showing average word happiness for geotagged tweets in all US states collected during the calendar year 2011. The happiest 5 states, in order, are: Hawaii, Maine, Nevada, Utah and Vermont. The saddest 5 states, in order, are: Louisiana, Mississippi, Maryland, Delaware and Georgia.
Le papier sur arxiv : http://arxiv.org/pdf/1302.3299.pdf The Geography of Happiness: Connecting Twitter sentiment and expression, demographics, and objective characteristics of place, par 5 auteurs de l’Université du Vermont