Critics have derided the 140-character messages posted daily on Twitter as trivialities. Yet to researchers, the popular social media site presents a rich trove of data. Barbara Poblete and her colleagues at Yahoo Research in Santiago analyzed tweets in the wake of February’s Chilean earthquake to learn how rumors propagate online. They found that people used Twitter to sort truth from falsehoods. Poblete’s group saw that 62 percent of tweets with earthquake-linked keywords from users in the Santiago time zone questioned or denied rumors that later turned out to be false. By comparison, when it came to confirmed truths, just 2 percent of tweets questioned them, and 0.3 percent were denials. Other researchers have used Twitter to track mood changes across the U.S. Alan Mislove, a computer scientist at Northeastern University, and Sune Lehmann, a Harvard University physicist, analyzed tweets that used words psychologists rated for emotional heft, such as “triumphant” and “suicide.” Their preliminary findings revealed that early mornings, late evenings and weekends rated the highest for happiness and that, unsurprisingly, the West Coast was happier than the East Coast.

The group now hopes to use Twitter to track changes in political climate. “Twitter is designed to be open, so we can look at content without violating anyone’s privacy,” Lehmann says. In April, Twitter announced it would donate its public tweet history to the Library of Congress. Researchers, at least, will be in a good mood about that.

This article was originally published with the title There's Wisdom in Those Tweets.

2 Comments

1. 1. twbeosl 09:48 AM 10/7/10

   The research seems to suggest there is wisdom ACROSS the tweets rather than IN them per se.
   
   This would seem to correlate with the "wisdom of the crowds/swarm" phenomenon that has been documented and more or less proven mathematically for dealing with simple problems, such as the number of jellybeans in a jar or the height of a building.
   
   But with Twitter, and other social networks like it, the more important question for the average user may be whether said user actually sees this larger group wisdom, or whether, since one's followers and followees are self-selected, the result on the ground, as it were, is more of an echo-chamber effect of that person's own biases.
   

   Reply | Report Abuse | Link to this

2. 2. bpoblete 02:04 PM 10/8/10

   The full research paper of the earthquake in Chile study "Twitter Under Crisis: Can we trust what we RT?" by Marcelo Mendoza, Barbara Poblete and Carlos Castillo can be found at http://research.yahoo.com/node/3255

   Reply | Report Abuse | Link to this