Monday, November 19, 2007

Signal-to-noise

This is an interesting bit of technology I read about in GMSV. And check out the random examples of stupidity on their website. Now if they could only get this technology to apply to the real world.


Momma says stupid is as stupid does

It sounds like something that anyone who’s run a forum or a blog has longed for at one time or another:
a stupidity filter — something that would scan incoming comments and, when necessary, prompt the would-be contributor with something like this: “This comment is more or less unintelligible. Please try to restate it.” What a boon it would be for the signal-to-noise ratio, but given the scope and diversity of the adversary, where would one begin?

Well, the folks behind
StupidFilter are trying to take a shot at it. As explained on the Web site, “The solution we’re creating is simple: an open-source filter software that can detect rampant stupidity in written English. This will be accomplished with weighted Bayesian analysis and some rules-based processing, similar to spam detection engines. The primary challenge inherent in our task is that stupidity is not a binary distinction, but rather a matter of degree. To this end, we’re collecting a ranked corpus of stupid text, gleaned from user comments on public websites and ranked on a five-point scale.”

Since it isn’t true artificial intelligence and has no awareness of context, humor, irony or sarcasm, the immediate goal of the filter is relatively modest: to detect “comments with too much or too little capitalization, too many text-message abbreviations, excessive use of ‘LOL,’ exclamation points, and so on.” And as
the FAQ explains, “We consider the StupidFilter’s irony-ignorance to be a feature, insofar as even if an allegedly-smart person makes a short, stupid comment, their smartness doesn’t make the comment any less stupid.”

Clearly, these folks have an immense task ahead of them. Any suggestions for other red flags that would clearly identify a comment as stupid?

No comments: