A computer program that can detect sarcasm online

Illustration for article titled A computer program that can detect sarcasm online

Sometimes it's hard to tell when somebody online is complimenting you or being a douche. That's why a group of Israeli researchers invented the "Semi-supervised Algorithm for Sarcasm Identification," a program that recognizes sarcastic statements with 77 percent precision.

That's right - a computer program is almost as good as you are at recognizing sarcastic comments.

According to Popular Science:

To create such an algorithm, the team scanned 66,000 Amazon.com product reviews, with three different human annotators tagging sentences for sarcasm. The team then identified certain sarcastic patterns that emerged in the reviews and created a classification algorithm that puts each statement into a sarcastic class . . . The algorithms were then trained on that seed set of 80 sentences from the collection of reviews. These annotated sentences helped the algorithm learn what sorts of words and patterns distinguish sarcastic remarks – those that mean the opposite of what they literally convey, or that convey a sentiment inconsistent with the literal reading.


So why would these researchers go through all this trouble to identify sarcasm? No, it's not to help humanity reach enlightenment. They're trying to help marketing companies figure out what people are saying about their products online. Imagine that "online focus group" meeting: "Well, everybody is saying positive things about your new book. Unfortunately, our software has determined that at least 80 percent of those comments were sarcastic." Hopefully there will be some altruistic applications for this program too - like helping the humor-impaired to figure out that no, I don't really think The Human Centipede needs a sequel about tardigrades.

Even if the sarcasm detector can be used for the forces of goodness, it saddens me to think of a computer learning about human satire from Amazon reviews, I suppose every AI has to start somewhere.

via Popular Science [via Slashdot, which has the full scientific paper]

Share This Story

Get our `newsletter`


Chris Braak

Now, we must construct a NEW KIND of sarcasm, to prevent the machines from understanding our secret thoughts.