Supercomputers That Predict Human Behavior

There is factual news and then there is news presentation and nuances and even reporting tone. Simply through the choice of words, a news article intended to convey factual information, becomes a commentary and persuasion piece. For example, 2 stories on the same subject may have headlines that read, “Prime Minister Fights to Maintain Party Control” versus “Prime Minister’s Party Expected to Win Election”.

The reason this concept of journalistic influence is discussed first is due to the fact that the BBC announced a USA university computing group is developing a supercomputer that can predict world events by relying on published news stories. The computer is fed information from government websites including BBC Monitoring and Open Source Centre in addition to newspaper articles and online news outlet reports published in a variety of sources.

These two news monitoring agencies are fascinating because they collect multiple information sources on the same topic and then report the story and how the story is being told around the world. In other words, a whole new level of analysis is added to news reporting. This laid the foundation for the supercomputer concept.

The SGI Altix Supercomputer is named Nautilus and is programmed to identify a news report’s mood as positive or negative and to analyze the actual event. The computer then prepares graphs and statistical data on information interrelationships and makes predictions. The growing sense is that a computer like this may be able to do a better job of predicting future world events than traditional government intelligence. The computer can process more and faster data and does not have built-in bias. Evaluation of mood by the computer is called “automated sentiment mining” and relies on the descriptive words chosen like “nice” or “horrible”.

So far the computer has charted Egyptian and Libyan revolutions, the Arab Spring, possible Osama Bin Laden locations (before his death), the turning sentiment against Egyptian Mubarak despite President Obama’s supportive words and others. To date the analysis has been conducted on past events, but real time analysis will be the next stage.

The question in my mind is how the computer will handle reporting bias. One news agency may call a political uprising “necessary and enlightening” while another may call the same uprising as “tragic and unnecessary.”  Perhaps the sheer quantity of information creates statistical certainty. However, there will always be a need for sensible government leaders who know that computer analysis is only the first step in decision making. It is always possible the supercomputer may not be so “super” and will get it wrong.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s