Researchers find flaws in using source reputation for training automatic misinformation detection algorithms

Researchers at Rutgers University have found a major flaw in the way that algorithms designed to detect “fake news” evaluate the credibility of online news stories.

This post was originally published on this website.