By Mike King
A day doesn’t seem to go by without another new development in social media monitoring appearing on the scene. One of the latest to gain publicity is a project at Northeastern University in Boston that uses Twitter messages to measure the mood across America. The "pulse of the nation" project monitors the Twitter feeds of 300 million Americans to produce a real time mood map of the country.
The analysis is based on interpreting keywords in 140 character messages to determine the mood of the sender. They’ll be no way to measure irony, sarcasm or, I expect context, so it is probably about as accurate as most other automated social media monitoring tools. Still, the findings so far reveal that the West coast is the place to be if you want to be happy (apparently, East coast people are just not as jolly), and that the nation’s happiness level peaks on a Sunday morning and troughs on a Thursday evening. No Monday morning blues in the USA it would appear.
The analysis is carried out using ANEW ratings (Affective Norms for English Words) where words are given emotional ratings in terms of pleasure, arousal and dominance. I don’t pretend to understand it fully, but words like ‘sick’, ‘bad’ and ‘wicked’ must prove a challenge.