Supercomputers can predict social unrest

In the future supercomputers will be able to predict social unrest events such as waves of protests in Egypt or civil war in Libya.

Discovery said Kaley Leetaru, a digital media analyst at the University of Illinois in the US, gathered a database of 100 million articles from 1979 to 2011. These articles were taken from the repository. data of US and British intelligence agencies.

By analyzing the content of the articles and their tone, Leetaru discovered that some types of articles frequently appeared during periods of social instability. For example, in Egypt, articles of negative tone about former President Hosni Mubarak gradually increased until he resigned.

Anti-Egyptian people protested in Cairo on January 25. (Photo: AP)

The stability of tone in articles can also help computers predict social situation. For example, despite the increasingly fierce criticism of the media, the Saudi government still does not face the risk of collapse because negative government articles still appear regularly with high frequency in many. last year.

Supercomputers can also detect Osama bin Laden hiding in Pakistan by examining the frequency of his name in articles relating to Pakistan. It then gathers those articles and lists the mentioned places. By connecting cities with straight lines and analyzing them, supercomputers concluded that bin Laden was hiding in northern Pakistan.

Of course, Leetaru's technology cannot predict events with absolute accuracy. That is like the weather forecasting technology in the early stages. Initially meteorological experts only predicted with average accuracy, but today the level of accuracy is large enough for people to make decisions based on weather forecast information. Supercomputers cannot guess the actions of individuals, but can guess the response of a community or society.