Even if we are incapable of following more than 150 friends on twitter, our computer overlords can track the humdrum ongoings of hundreds of millions of people with ease. What's more, new research suggests that seemingly inconsequential tweets (in large enough numbers) can have seriously far-reaching implications for people who play the stock market.


Anyone who has wasted an afternoon playing around with Google Trends (or the more advanced Google Insights) can appreciate the wealth of information hidden in web-based data; applications like these, and social networking outlets like twitter and facebook, not only allow us to monitor the real-time interests of hundreds of millions of people — they invite us to make predictions about how those interests may impact the future.

Not surprisingly, a number of people have welcomed this invitation with alacrity. In 2009, Hal Varian, Google's chief economist, published a paper (PDF) that demonstrates how web-based data can be used to forecast information ranging from travel behavior to fluctuations in automobile sales. More recently, in a paper published in the March issue of the Journal of Computational Science, Dr. Johan Bollen, a professor at Indiana University School of Informatics and Computing, describes a computer algorithm that uses the collective mood of the United States (determined by analyzing tweets for emotional content) to predict swings in the Dow Jones Industrial Average.

As fascinating as these forecasting methods are, they're far from perfect. How, for example, does a computer go about interpreting a sarcastic tweet? Will this power be used for evil or good? Probably both. But don't let hypotheticals like these distract you from the most important take home message: someone, somewhere, cares about how totally awesome your lunch was, even if that someone is a computer.


Via The Economist
Top image via