Happy Monday, freaks!
So, if you're reading this, than it is safe to assume you have survived all the doom and gloom prophecies for the tenth anniversary of 9/11. And for that, I salute you!
In the vein of prophecy, check out this story today from the great news cannon across the pond.
From BBC: Supercomputer predicts revolution
As the story goes, a Kalev Leetaru of the University of Illinois Institute for Computing in the Humanities, Arts and Social Science has unveiled a new software program which he claims was able to predict the hot spots of the Arab Spring, as well as provide a surprisingly narrow search radius for Osama bin Laden before his death. Of course, all of these predictions were handled in hindsight, so one has to wonder how much the facts are being distorted to support the end result here.
Leetaru's program works by assimilating millions of media articles, and then dissecting them into two separate variables: mood, and location. Location is simple enough, as the computer simply takes any locations named in the article and translates them into coordinates to be plotted on a graph. The enumeration of mood is slightly more tricky. Defining mood levels, or "automated sentiment mining", was performed by analyzing the articles for key words, such as "terror", "horrific", or "nice", and then in a flash of computer wizardry, likely involving gremlins with flame-throwers, stamps a mood definition to pair with the location of the article.
Using these information sets, Leetaru was able to produce a graph over time displaying what he claims to be an accurate predictor of the mood of a given location. Indeed, the graph created for Egypt not only shows the fall of Mubarek, but also several other high tension periods in Egypt's recent history. It was able to do the same with Libya, and was also able to predict bin Laden's location within 200km of his lair in Pakistan.
The proof is in the pudding, as they say, and it remains to be seen whether Leetaru's expensive computer magic will be worth its salt when it comes time to predict a revolution that hasn't happened yet. But for those of us who have so little faith in the validity of anything that comes through the pipe-lines of the international media, than it would appear that Mr. Leetaru and his device are already beginning from a losing position.
No more than the rest of us, though.
Besides, I can't be the only person who feels that we might be allowing the computers to do a little bit too much of our thinking these days. Do we really need to take that extra step to have a machine predict when people have been pushed to the point of violent rebellion in the streets? Seems to me a decidedly inhumane response to such a humanitarian problem.
Besides, God Help Us if the machines were to go down.
No comments:
Post a Comment