By Evan Ackerman
If you spend any time at all on the internet, you know just how full of crap a lot of it is. I blame most of that on Web 2.0, since before that, people had to invest at least some modicum of effort to make a public spectacle of themselves. The StupidFilter Project is stupidly simple: it’s designed to be “open-source filter software that can detect rampant stupidity in written English.” From the sound of things, it’ll either work server-side (preventing users from, say, posting comments on a forum that contain excessive misspellings, exclamation points, use of text abbreviations, or suffer from terrible spelling or no punctuation) or client-side (quietly replacing “OMGLOLZ!!!!!!!!111” with “!,” for example). The big question is, of course, exactly what qualifies as stupid. After the jump, read what the developers have to say on the matter:
Since we’re trying to build a detailed database that serves as a very verbose example of What Not To Do, we look for comments whose prose style we can point to and say, “I don’t even have to understand the content of this comment to know that it’s stupid,” — based on the gross prose style alone, its stupidity is self-evident. It is then useful as an example for our parser to integrate into its database of stupidity. …We grade stupidity on a scale of 1 to 5. Someone might get a 1 or 2 for a comment that used no punctuation, whereas a comment consisting of nothing but text message abbreviations with a dash of LOLLLLL thrown in for good measure would probably rate a solid 4 or 5. There is a certain amount of subjectivity, and our software is aware of that; scoring will be normalized to eliminate excessively generous or harsh estimations of stupidity.
For a random sample of what StupidFilter is designed to catch and how the rating system works, click here. The sample may contain offensive language, but most likely only contains offensive spelling and/or grammar. An alpha of the source code is scheduled to be released in December; OMG I tottaly cant wait.