That has nothing to do with the kind of 'fake news' being discussed. There's nothing 'fake' about reporting "Government claims WMDs in Iraq" while elaborating on what evidence the government has or claims to have, even if it turns out there were no WMDs.
The fake news is question comes from sites that just make up random crap to get page views, not websites that are making a good faith effort to report on what is real.
You don't get to say a propaganda campaign that ruined a country and killed hundreds of thousands wasn't fake news. It was - complete and utter fabrication on an industrial scale, and it had devastating effects.
Your problem is that you're interpreting it in terms of "is wrong/is right" instead of "deliberately lying/trying to be accurate". Perhaps news publications should've been more skeptical, but they themselves weren't trying to make things up.
Officials supplied the false information which served as the raw material the media's fake news was produced with.
This is a famous quote from Judith Miller and Michael Gordon, from an article that includes many hawkish quotes from anonymous sources.
"Mr. Hussein's dogged insistence on pursuing his nuclear ambitions, along with what defectors described in interviews as Iraq's push to improve and expand Baghdad's chemical and biological arsenals, have brought Iraq and the United States to the brink of war."
http://www.nytimes.com/2002/09/08/world/threats-responses-ir...
Here's a quote from an opinion piece by William Safire:
"Two terrorist-sponsoring nations are racing to acquire nuclear weapons. One is Iraq, whose scientists already have the know-how..."
http://www.nytimes.com/2002/01/10/opinion/arafat-s-implausib...
Is this fake news? I'd say so. But, I have an even lesser criterion for fake news -- I think the most insidious type of fake news doesn't necessarily tell a single lie, but rather selectively reports the truth.
I am also 100% certain that it's impossible to filter out "fake news", whatever your definition is, without suppressing some true but fringe information.
The politicians involved knew that their claims were not backed up by facts.
They arguably should have been more skeptical but it's very different from the situation where some place like Breitbart runs a propaganda piece which is easily debunked using public information, like the fake electoral map they printed which flipped entire states to maintain the “coastal elites” narrative they're selling.
The mainstream media has some serious blemishes from that era (Valerie Plame comes to mind) but they have editors, publish corrections, etc. and it's a monstrous exaggeration to try to equate the two.
The paid protesters story is a great example: the campaign benefited from and recirculated rumors about that which started on random blogs, conspiracy sites like InfoWars, and outright fakes like abcnews.com.co throughout the election. The guy who tweeted that photo might have been a Republican acting in what he believed was good faith but he was primed to think of that as likely by the prior months of unfounded accusations and as soon as he shared that photo, it was widely circulated by supporters including Corey Lewandowski, who was Trump's campaign manager before becoming CNN's commenter representing the campaign.
Plus, is it really a good idea to suppress a meta-discussion about propaganda and false information? Those "fake news" articles and their corresponding beliefs will still exist, even if Facebook manages to filter them out, but proponents of those ideas will either never see critical discussion or they'll just move to other platforms. It seems like a great way to further fracture our already divided society.
You're getting hung up on the term 'fake news'. In this context this term has a specific meaning, different than what you're referring to. 'Fake News' are sites with content that is purposely false (though it could plausibly be true) that is also usually outrage-inducing. At the same time they attempt to pass themselves off as credible news organizations. Their business model involves getting their fake content to go viral, increasing views and their advertising revenue.
There's several levels to "truth".
What happened.
What an eyewitness/machine reported happened
Whether the witness maliciously lied/mistakenly reported the event
Was all of this reported faithfully
Separately was reporting done to verify the facts.
And do note, we've (coders) created the world where most newspapers have already gone out of business, and they have to in turn compete with everything and everyone for what remains of human attention spans.
So the new question I would append to that list is - does it even matter anymore ?
ProPublica for example tries very hard to do good in depth reporting. There have been reams of leaks with more data than what most people know what to do with.
This has to compete for human attention bandwidth against documents/content designed to press buttons, elicit emotions and thus propagate (generating more page views, and so revenue)
My point is simply that truth is a narrow view on the scope of the problem.
In a year or two this will die down and a new generation of coders will parrot "move fast and break things."
The larger system at play her must be front and center.
Ideally people would have more time to reflect on their own - see more news that causes dissonance - hopefully causing people to check their opinions and beliefs.
And of course the tool to Achieve this should itself not create a worse problem
And every situation is even more peculiar. This is a futile exercise.
There's no need to talk technical about graph nodes and machine learning. It's very simple - the fake news labels are, and will always be biased in favor of the liberal/progressive/Democrat agenda. Objective news is a lie. No one wants to hear it. We want echo chambers that make us feel good. The labels will shift to accommodate this desire.
Otherwise you would see the US population asking themselves why on earth are their armed forces so deep involved in Syria, just short of military invasion. They get spoon-fed that the country is ruled by a dictator and their population needs to be freed, but so is Saudi Arabia with a far worse regime where people get their heads cut on monthly basis, which are praised as critical ally in the region.
"Fake news" is a way to block the things deemed as dissident from the main stream, effectively blocking the room for logical argument and exposing the facts.
There was no mention of machine learning (no predictions of any kind) - the idea is to keep track of history of raw data (labelling and it's changes) and report on "reliability score ..." (how many times labelling itself was false).
The thing is it doesn't need to be biased at all. You just keep all data for all sources. There's no "objective truth" per se, there are only true/false labels according to source.
Of course you can create your own "beyond reasonable doubt by xyz" or "government xyz pov" source and assume that this is going to be your objective truth reference. But, just like any other sources, your source will just have reliability rating in context of any other source.
Labels will probably have to change retrospectively. Maybe they should also include "according to ..." reference? I can imagine that some statements will be flagged as truth by one side and as fake by the other.
If you play this game further with creating graph of dependencies and recorded switches from truth to fake you can even get "accuracy score" for source reference - I wonder how CIA as source ref would score in this game - hopefully not somewhere in the area of "compulsive liar"?