Social networking is a big deal, and not just for smart phone-addicted teens and reconnecting with long-lost friends. Twitter has grown into an incredibly useful way to disseminate information, and many reputable institutions, including PLoS and PLoS ONE, use it to share news and updates with thousands of people across the world.
Researchers are beginning to key into the movement too, not just as users, but also as active investigators of the dynamics and utility of these new technologies as they emerge and grow. One of the amazing effects of these social networking sites is the wealth of data they can provide, and now scientists are taking advantage of the huge amount of public content from sites like Twitter, Facebook, and even Wikipedia to answer all kinds of new questions.
For example, investigators showed that Twitter is a useful tool for tracking H1N1 here and here; that content disputes in Wikipedia reflect geopolitical instability; and that virtual social networks can create collective emotional states.
Today, we added to this growing body of work with a report of a Twitter-based “hedonometer” that can be used to quantify the societal happiness of large populations. The authors used an amazing library of 46 billion words from nearly 4.6 billion tweets posted over almost three years by over 63 million unique users – a collection that would have been essentially impossible to obtain without the Twitter-verse.
The paper reports various trends in happiness – people are happier on the weekend, and the word “Christmas” is associated with high happiness levels, as opposed to “flu” and “Iraq,” which rank at the bottom – but the real advancement of the paper is its quantitative approach to the huge Twitter-based dataset. While scientists now have access to these huge datasets, they must first face the challenge of classifying and organizing the huge amounts of social information so they can conduct meaningful research into areas that have not previously been explored.
Image courtesy of La Fabrique de Blogs