New Scientist (11/17/12) Jim Giles
Georgia Institute of Technology professor Jacob Eisenstein and colleagues are using social networks to examine the evolution of language. The researchers collected 30 million Tweets sent from U.S. locations between December 2009 and May 2011, and built a mathematical model to capture the flow of new words between cities. The model found that new words tend to take life in cities with large African American populations before spreading more widely. Moreover, cities that are economically and ethnically similar are more likely to share new words. "Their results indicate that birds of a feather tweet together," says University of Groningen linguist John Nerbonne. The researchers are working on a more detailed analysis that could potentially reveal which cities are most influential. Eisenstein also wants to know whether neologisms spread more quickly because of Twitter and other social networks. An analysis of many types of data, including blog posts and Facebook entries, would enable the team to study whether social media is accelerating the evolution of language more generally.
http://www.newscientist.com/Georgia Institute of Technology professor Jacob Eisenstein and colleagues are using social networks to examine the evolution of language. The researchers collected 30 million Tweets sent from U.S. locations between December 2009 and May 2011, and built a mathematical model to capture the flow of new words between cities. The model found that new words tend to take life in cities with large African American populations before spreading more widely. Moreover, cities that are economically and ethnically similar are more likely to share new words. "Their results indicate that birds of a feather tweet together," says University of Groningen linguist John Nerbonne. The researchers are working on a more detailed analysis that could potentially reveal which cities are most influential. Eisenstein also wants to know whether neologisms spread more quickly because of Twitter and other social networks. An analysis of many types of data, including blog posts and Facebook entries, would enable the team to study whether social media is accelerating the evolution of language more generally.
