快猫短视频

2020 vision: The crystal ball internet

Sentiments expressed in the torrent of blog posts, tweets and Facebook updates offer a powerful way to predict the future
Text mining can help forecast stock market movements
Text mining can help forecast stock market movements
(Image: Yoshikazu Tsuno/AFP/Getty Images)

Read more:Seven technologies to disrupt the next decade

For around 20 years, starting in the 1980s, the sought predictions from people considered knowledgeable. His experts, 280 of them, were the kind of folk who, in their work as TV pundits or government advisors, opined on matters such as the rise of China or security in the Middle East. As time passed, he checked their forecasts. The results were dismal. 鈥淗uman beings who spend their lives studying the state of the world鈥 are poorer forecasters than dart-throwing monkeys,鈥 wrote .

Not so for a powerful new method of forecasting called 鈥渢ext mining鈥. It draws on the vast amount of data available online. By sampling the sentiments expressed in the torrent of blog posts, tweets and Facebook updates, you can gain unprecedented insights into the mood of the world and use it to predict what is to come.

Researchers have already developed powerful enough to improve forecasts of . Others have used Google search queries as a forecasting tool. Many searches for certain job-related terms, for example, indicates that unemployment is rising.

That鈥檚 just the beginning. Several companies are now archiving whole swathes of the internet in a bid to create more powerful forecasting. WiseWindow, based in Irvine, California, claims to monitor opinions expressed by over 77 million people on Facebook and other social media sites. The firm mines the data for clues to consumer sentiment and emerging trends, which companies buy in the hope of getting an edge over competitors. Such companies鈥 forecasts have not been tested in depth, at least not publicly. But if these endeavours prove as successful as the studies based on tweets and search terms, forecasting could get a lot better.

This could be good news for all of us. If governments can get a better handle on economic trends, for example, they might be able to nip recessions in the bud. Could an early-warning system based on social media have helped to prevent the financial crisis of 2008? It鈥檚 too early to make such a bold claim, but the work done so far suggests that the idea is not fanciful.

Yet there is also a down side. The blog posts and tweets in which we share our thoughts and feelings are all now a target for advertisers. We are all part of a vast market research project, whether we like it or not.

Lexicon of tomorrow: SENTIMENT ANALYSIS

-苍辞耻苍听听听听听听听听听听听听

A method for gauging the real-time emotions and opinions of a demographic or region using text mining. The political analyst performed a sentiment analysis of tweets in Virginia ahead of the election