Reuters Built An Algorithm That Can Identify Real News On Twitter (popsci.com) 122
Reuters has built an algorithm called News Tracer that flags and verifies breaking news on Twitter. The algorithm weeds through all 500 million tweets that are posted on a daily basis to "sort real news from spam, nonsense, ads, and noise," writes Corinne Iozzio via Popular Science: In development since 2014, reports the Columbia Journalism Review, News Tracer's work starts by identifying clusters of tweets that are topically similar. Politics goes with politics; sports with sports; and so on. The system then uses language-processing to produce a coherent summary of each cluster. What differentiates News Tracer from other popular monitoring tools, is that it was built to think like a reporter. That virtual mindset takes 40 factors into account, according to Harvard's NiemanLab. It uses information like the location and status of the original poster (e.g. is she verified?) and how the news is spreading to establish a "credibility" rating for the news item in question. The system also does a kind of cross-check against sources that reporters have identified as reliable, and uses that initial network to identify other potentially reliable sources. News Tracer can also tell the difference between a trending hashtag and real news. The mix of data points News Tracer takes into account means it works best with actual, physical events -- crashes, protests, bombings -- as opposed to the he-said-she-said that can dominate news cycles.
Re: (Score:2)
Nice try HUMA ABEDIN.
Re: Sad to see Trump has forced... (Score:1)
He's already creating jobs.
Re: (Score:2)
By essentially blackmailing a defense contractor. Sure, it works, but you can't do this large scale. And threatening to take away some business with government if they don't do what they're told elsewhere is what many leftist/socialist governments do to force businesses to fall in line, which seems somewhat ironic here.
"Built to think like a reporter"... (Score:5, Funny)
Does it vote Democrat, as well?
Re: (Score:1)
Re: (Score:2)
Re: (Score:2, Insightful)
Yeah, if it's from Reuters and "thinks like a reporter" then the algorithm will be quite simple
IF (newsSource = conservative)
THEN (fakeNews = true;)
ELSE IF (newsSource = liberal;) THEN (fakeNews = false;)
ELSE {SEND(story,DNC,vetting);}
Re: (Score:1)
I agree, since the DNC has adopted Nixonian tactics, then they should arrested just like those assholes were.
Re: (Score:2)
Re: (Score:2)
Concentration Camps:
Japanese, German, and Italian Americans during WW2.
and Killed Them:
only the ones that enlisted in the military (Cannon Fodder [wikipedia.org]).
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Very likely....does not vote. But it sure as heck campaigns Democrat.
Amazing (Score:1)
I bet all their news is the "real" news!
The source code (Score:5, Insightful)
if (article.agenda() != ourAgenda) article.fake = true
Re: (Score:2)
Or you can simplify the forced boolean state in most languages for not-quite-boolean returns by using a double-not !! before the function name!
Oh boy, the media is not bias (Score:2)
What's the old joke? Reality has a liberal bias.
Re: (Score:2)
Ha, you must be a time-traveler from the pre-SJW leftist era. These days the apeshit coming from the SJW left makes the old A.M. radio batshit coming from the right look positively sane.
It's a pretty bizarro world where liberals are now the ones screaming for banning free speech and bullying their opponents into silence. They've even managed to one-up conservatives on their conspiracy theories. I remember laughing after Obama's election when pawn shops were reporting a run on gun-buying from gun nuts conv
Re: (Score:3)
Re: Oh boy, the media is not bias (Score:2)
No he hasn't.
WHAT GOT US TRUMP (Score:2)
Was having the Democrat party engage in the largest mass voter disenfranchisement and fraud in U.S. history in order to push the most corrupt and corporate connected politician to ever run for POTUS.
THAT is what got us Trump.
Re: (Score:2)
Actually its:
if (Trump.isPissed()) article.real = true
EXACTLY!!! (Score:2)
This pretty much sums up the Al-Gore-ithm.
Re: (Score:2)
Also blacklist any article which starts with a number in the title. This will kill about 95% of buzzfeed bullshit right off the bat.
Re: (Score:2)
I don't think that would suffice, though it's certainly a reasonable weighing consideration. And what they specified as the technique wouldn't suffice, either, though it would reduce the fake news considerably. And invite a "fake news" arms race.
At some point verification requires somebody you reasonably trust actually going and checking. I suppose a video might count, but you need to consider that every video is going to provide you with a biased view, selected by the angles from which observation happe
No (Score:3, Funny)
Reuters Built An Algorithm That Can Identify Real News On Twitter
Lies.
I hope it flagged itself.
The litmus test (Score:5, Insightful)
Does anything on Huff, and WaPo pass?
You know the funniest thing about everyone talking about "fake news"? They make it look like it's only a conservative rag problem. People's memories are so razor short these days, they've already forgotten that The Rolling Stone published literal, fake "news" about a campus rape story, ruined peoples lives, and were sued for 8 (reduced to 3) million dollars.
If people here were half as skeptical as they claim to be, they'd have no respect for conservative AND liberal "journalists." Science demands proof. It doesn't care if the lack-of-data is coming from people you like.
Re: (Score:2)
Science demands falsifiability.
FTFY
Re: (Score:2)
Re: (Score:2)
> I remember that Rolling Stone story, it wasn't fake news but it was bad journalism.
How, exactly, do you distinguish those?
And if anything, doesn't that imply that "bad journalism" is more dangerous because people are more likely to trust it when it comes from a supposedly-reputable source?
Re: (Score:2)
Re: (Score:3)
So when CNN has a lawyer employed by them tell us false statements about the law, what does that count as?
Chris Cuomo on CNN -
https://www.youtube.com/watch?v=_X16_KzX1vE [youtube.com]
An explanation of how badly wrong he is:
https://popehat.com/2016/10/17/no-it-is-not-illegal-to-read-wikileaks/ [popehat.com]
Chris' bio on Wikipedia showing he's a licensed attorney:
https://en.wikipedia.org/w/index.php?title=Chris_Cuomo&oldid=752709245 [wikipedia.org]
Also, we have a motive in that they rigged the debates with the DNC in DKIM-validated emails signed by
Re: (Score:2)
Re: (Score:2)
Do you understand the difference between an opinion that turns out to be incorrect as a matter or law, and just making up completely fake clickbait to enrich yourself at the expense of democracy?
To put it another way, while occasionally CNN might get flagged by this, other sources would have 90%+ of their output flagged.
Re: (Score:2)
You assume, without evidence, that people actually believe clickbait. No, what people actually believe is when people present evidence. When they do real investigations, which have all but stopped for budgetary reasons.
And no, this wasn't an understandable wrong opinion. That was horribly, badly wrong by someone who should have known better. Given that it tended to cover up CNN's own misdeeds, I'm not having an easy time writing that off as a mere mistake. As someone else put it, "false exculpatory sta [youtube.com]
Re: (Score:2)
Also, what about CNN interviewing its own cameraman?
https://www.youtube.com/watch?v=0_W5cDjy3uU [youtube.com]
Or editing what people said to convey the opposite message?
https://www.youtube.com/watch?v=S-8Cn6boqcA [youtube.com]
Are these all accidents?
Anyhow, my point would be to look at the actual facts in a story (if any) and totally filter out the opinion and predictions. The source of facts doesn't matter, what really matters is whether they're verifiable or not. Trying to rebut facts with opinions doesn't work. It just makes that
By this definition (Score:2)
We can label CNN, ABC, Fox, MSNBC, HuffPo pretty much ALL fake news outlets.
Re: (Score:3)
Not really. There's only confirmation bias. It can be on the part of the reader or on the part of the journalist but it's just as bad either way. People will even take OBVIOUS satire sites and take them seriously so long as it fits their internal narrative.
"Journalists" do this too. They will ignore stories that don't fit their narrative. They will rush to judgement when it suits their narrative.
Re: (Score:2)
I think the difference is (Score:1)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Really bad photoshop, you covered up half the image with a white text background.
Re: (Score:2)
That image appears to be fake. The article isn't on their web site and no one seems to have archived it. Beyond that, it's a rather poor Photoshop, MS Paint quality really.
Re: (Score:2)
While I agree with everything you've said, you're making false equivalences... One (huge) mistake doesn't turn a legit news organization into a supermarket tabloid, just as a few lies on one side doesn't balance out a voluminous blatant and continuous intentional disinformation campaign on the other side.
THAT is a perfectly valid reason why discussion on the topic tends to be one-sided, even if problems on the other side need to be resolved as well.
That was not fake news again (Score:4, Insightful)
Re: (Score:2)
Fake news are hoax news made especially to lie to people.
That's all well and good, but there's a spectrum between hoaxes, spin, and unbiased (as is reasonably possible) factual reporting. What the mainstream media is trying to do is conflate "fake news" with alternative, right-wing news sites that put their own spin on the news, but it isn't "fake".
As should be obvious by now, the mainstream news has a left-wing bias and apply their own spin to stories, sometimes more blatantly than others. And there are plenty of blatantly left-wing sites to correspond with righ
No (Score:2)
"Reuters Built An Algorithm That Can Identify Real News On Twitter"
No they haven't. Ask me how I know.
How news spreads? (Score:2)
[...] how the news is spreading to establish a "credibility" rating for the news item in question.
I wonder if it takes into account the old saw about how "A lie can travel around the world while the truth is lacing up it's boots." [quoteinvestigator.com]
Algorithms != Implementations (Score:2)
Reuters did not build an algorithm. They devised an algorithm and then built a system based on that algorithm.
Algorithms are methods... processes... ways of doing things. Algorithms are not implementations. Algorithms are the conceptual steps, not the manifestation of those steps.
Re: (Score:2)
You're using a very narrow definition of "build".
Build (verb): 7) "to form or construct a plan, system of thought, etc"
http://www.dictionary.com/brow... [dictionary.com]
Completely valid use of the word.
Re: (Score:2)
Re: (Score:2)
Um whatever dude, you said this:
That's what I replied to.
TFA says "Reuters has built an algorithm called News Tracer that flags and verifies breaking news on Twitter."
Again, perfectly valid language here. This is not "misuse" of either "algorithm" or "build". Your complaint is asinine.
Official liars (Score:2)
Re: (Score:2, Troll)
true [dailymail.co.uk]
So... (Score:2)
For propaganda definitions of "real news". (Score:2)
If this was as good as it supposed to be, it'd put nearly everyone at Reuters out of a job. Their primary product is fake news.
The only thing this will do is use natural language to identify narrative-breaking articles before they can gain traction with the public.
Still searching (Score:2)
The algorithm weeds through all 500 million tweets that are posted on a daily basis to "sort real news from spam, nonsense, ads, and noise,"
Has it found any, yet?
Initial fake news story was so easy to spot (Score:1)
Real News Definition (Score:2)
What is the criteria? (Score:2)
For this to be of any use, the criteria needs to be posted publicly. For example, very little proof has been proffered regarding Russia's supposed involvement in hacking the DNC. And many say it looks more like forces inside the U.S. intelligence community.
So how does this Reuter's tool handle such a case? Claiming that a politically appointed security chief of the government states something, therefore it must be factual or true is a very poor measure. What if a Trump appointed security chief states som
Re: (Score:2)
Distinguish between fake and wrong. They don't mean the same thing.