Stuff We Want: A ‘Stupid’ Filter
They say spam exists because some people fall for it. By the same logic some people must love comments like OMG!!! ROFL!!! and the sort of idiotic nonsense that that seems particularly endemic to YouTube, but affects nearly every site to some degree. In fact, it’s possible that the same set of people are responsible for both problems — stupid people.
We’re sort of lucky here at Compiler, that majority of the comments posted here are well thought out and well written, but if you’ve got site where that isn’t the case, a possible solution is in the works. The StupidFilter Project is an attempt to filter comments based on, well, how stupid they are. As the project’s website says:
StupidFilter was conceived out of necessity. Too long have we suffered in silence under the tyranny of idiocy. In the beginning, the internet was a place where one could communicate intelligently with similarly erudite people. Then, Eternal September hit and we were lost in the noise. The advent of user-driven web content has compounded the matter yet further, straining our tolerance to the breaking point.
The solution, according to the folks behind the stupidity filter, is to employ a Bayesian filter (like spam filters) that clears out a cruft and only lets through comments of some substance.
On a practical level this is no simple task. As the project’s head Gabriel Ortiz, recently told Fortune Magazine, “smart people are often ironic,” and irony can look stupid to a computer.
But before you dismiss the stupidity filter as impossible (which it may well be) at least consider that the group does realize the complexity and inherent subtle nuances of linguistics that complicate their task.
For instance, as the Fortune piece points out, “the clueless tend to repeat consonants: ‘This video is amazinggggg!!!’ By comparison, says Ortiz, ‘when you repeat a vowel, you’re being sarcastic — ‘Yeaaaaaah.’ We’ll be using several different methods to try to mediate this.” We should point out that this isn’t true 100% of the time, but it also isn’t a bad place to start.
Of course, what would the system make of the people who will inevitably comment on this piece using various strings of l33t speak and exclamation points? Are they stupid or are they being clever?
One nice touch in StupidFilter’s proposal is that the system won’t just filter the comment out, it’ll offer the stupid a chance to revise and articulate their ideas in a more coherent manner. If they fail to do so, then they’ll be rejected.
Will it work? Probably not, but here’s hoping. And in the mean time would it be so hard to just filter out any comment that contains more than one exclamation point in a row?
