You choose, we deliver
If you are interested in this story, you might be interested in others from The Journal Gazette. Go to www.journalgazette.net/newsletter and pick the subjects you care most about. We'll deliver your customized daily news report at 3 a.m. Fort Wayne time, right to your email.

Science & Tech

  • Wireless industry makes anti-theft commitment
    A trade group for wireless providers said Tuesday that that nation’s biggest mobile device manufacturers and carriers will soon put anti-theft tools on the gadgets to try to deter rampant smartphone theft.
  • Study: Snack may avoid fight with spouse
    A quick candy bar may stave off more than hunger. It could prevent major fights between husbands and wives, at least if a new study that used voodoo dolls is right.
  • Uber meets local lookalikes in Asia taxi-app wars
    Riding on its startup success and flush with fresh capital, taxi-hailing smartphone app Uber is making a big push into Asia. There’s a twist, though: Instead of being the game-changing phenomena it was in the U.S.
Advertisement
File

Korean scientists' tool helps separate fact from fiction on Twitter

Bigfoot was finally discovered in 2009 - at least according to rumors circulating on Twitter. With misinformation rife on social media, users could do with a tool that can sift truth from fiction.

Now Sejeong Kwon and colleagues at the Korea Advanced Institute of Science and Technology have designed an artificial intelligence system that, they claim, does this correctly around 90 percent of the time. If built into social networks, it could help people avoid embarrassing retweets or reshares of false information.

The system analyzed language used in more than 100 rumors - some later confirmed, others unfounded - that went viral on Twitter over a period of 3 1/2 years. The researchers found that false rumors were far more likely to contain negative terms such as "no" or "not" than positive terms such as "like" or "love."

Being mentioned in "singleton" tweets - ones that were neither retweeted nor replied to - was another indicator of false content. Indeed, the best predictor that something was false was that it was tweeted separately by many users; accurate stories tended to have a few, widely retweeted sources.

Advertisement