EU project to build lie detector for social media
I am not sure how practical this is, technically, as I see it as necessary to record the first occurrence of any new trend and there has to be many, many first occurrences that never turn into a trend. Also, they would have to record the origin of each occurrence, somehow classify its truthiness, _then_ track how it trends. An interesting technical challenge, something I think would be cool to try, but there are so many sources of nonsense you would have to monitor them all (for instance, how many times has some story made up on the Onion wound up being taken seriously by some nimrod who couldn’t be bothered with actually checking the source?) and to do so even when the networks get clogged with people pinging one another back and forth on the very same nonsense.
I would love to see it successful, though I have to say since I started responding to my relatives who spam me with nonsense by sending them a link to Snopes I have got a lot less of that type of spam. It is possible that they are now checking Snopes, but perhaps they just leave me off their distro list.