Hot on HuffPost Tech:

See More Stories
AOL Tech

Algorithm Detects Sarcastic Tweets and Amazon Reviews

Sarcasm detected in amazon reviews
For those who often miss out on a virtual joke, scientists have found a way to help identify online sarcasm. According to Geekosystem, a team from the Institute of Computer Science at The Hebrew University in Jerusalem developed an algorithm that detects sarcastic sentences posted online. The Semi-Supervised Sarcasm Identification Algorithm (SASI) learns to detect sarcasm by studying sentences that humans tag as sarcastic.

In the team's recent study (.PDF), the algorithm correctly identified tongue-in-cheek tweets and Amazon product reviews 77-percent of the time. According to the research, 'The Da Vinci Code' novel, the Kindle and both Shure and Sony's noise-cancellation earphones are the most sarcastically reviewed products on Amazon. To nobody's surprise, the use of all-caps and exclamation marks were a good indication that somebody's being sarcastic. Maybe the team will reverse hack the algorithm to make YouTube's comments section all but obsolete. [From: Geekosystem]

Tags: algorithm, amazon, computer, funny, Hebrew University in Jerusalem, HebrewUniversityInJerusalem, humor, research, sarcasm, sasi, socialnetworking, study, top, twitter, web