Wow, what a breakthrough: researchers develop online sarcasm detector

No, I’m not being sarcastic, this is actually a pretty big deal, or at least the first step down a road towards something that is actually a pretty big deal, because if you’ve used one of the myriad social media sentiment trackers that have suddenly sprung up in the last year or so you’ll know how patchy they are. Particularly as sarcasm, irony, LOLS and other types of non direct speech are pretty much the default mode for a lot of Internet dialogue.

Researchers at Hebrew University have developed a machine algorithm that can recognise sarcasm in online postings. The algorithm was tested against 66,000 reviews for books and other products listed on Amazon, where it apparently achieved 77% precision in identifying sarcastic comments.

You can read the full paper on sarcasm here. However, if you just want to know more about sarcasm, irony and the like, then read DJ Enright’s brilliant essay on the subject – The Alluring Problem.

Leave your comments

Your email address will not be published. Required fields are marked *