Korean researchers, in cooperation with Microsoft Research Asia in China, have created technology to filter rumors from postings on social networks by investigating 100 cases of rumors and true stories spread on Twitter in the United States within three years.
The scientists looked at the numbers and used sociology and psychology theories to find out what a rumor looked like. They admitted it was difficult to classify rumors at the very early stage of their dissemination, but after time passed it became clearer what was rumor and what was truth.

The researchers explained that they had to look at the characteristics of rumors, like how long they spread for, who distribute them and which words they used. Soon the scientists could correctly identify 90% of unfounded claims as rumors. For example, among the big ones that kept appearing as a rumor were allegations that Barack Obama was a Muslim and illegally obtained his US citizenship.

Apparently, rumors spread over a long period, but truth can be widely spread in a short time, and then is rarely mentioned again. Besides, rumors are also circulated randomly by people who don’t know each other while the spread of ordinary information can be tracked along the networks of online acquaintances. It also turned out that rumors often contained words associated with assumption and doubt: “not sure”, “no idea if it works”, and “heard”.