Thursday, January 10, 2019

Detecting Fake News: Not As Difficult As You Might Believe - Unless We Reach "The Inversion"

Image may contain: text

No photo description available.

Fake news has become like a pandemic mind virus, and there's no sign it will abate soon, certainly so long as the Trumpies control the power center of the nation - and spew nonstop lies.  However, people will be glad to know it's not all a lost cause, there are ways to spot fake news and  not be suckered in by it. This is important,  especially recognizing the top fake news sources which appear to be Facebook and Twitter. In the graphs shown, for example, we see how two fake news stories were given new life on the child- level platform called Twitter. The favorite tool of Trump,  for obvious reasons, namely that he acts and writes at the level of a dyspeptic five year old.

The disturbing finding of the two graphs - actually quite depressing- is that the majority of Twitter users who discussed false rumors about the two named disasters, posted tweets that actually spread the rumors. This is exactly analogous to a mind virus infection spread from a single source by either unknowing or intentional human misinformation vectors.   Only a small fraction of these users sought verification or expressed doubt about the content.  This means as our nation gravitates more to this childish medium there is ever greater risk of becoming a nation of brainwashed sheeple. Or, paving the way for Civil war between the sheeple and normal, self aware, rationally- grounded citizens.

Some may say the solution, whether on Facebook or Twitter, is to scrub out bogus articles or block irresponsible tweeters from infesting the web with their nonsense.  But this is not the way to go because simply scrubbing BS articles or tweets will likely bring up a reactionary backlash, i.e. "What is this,  Nazi Germany?"    In the words of  Murphy Choy, a data analyst in Singapore, quoted in a Science News piece from August: "That would be like a totalitarian state. It would be very uncomfortable for all parties involved."

That means we have to train net users to become au fait with detecting fake news, and there are ways to do that. (Critical thinking practiced regularly also helps!)

A first step is provided by two computer scientists - Sibel Adali and Banjamin Horne - based at Renssalaer Polytechnic Institute. The pair analyzed 75 true articles from media outlets deemed most trustworthy as well as 75 false stories from sites on a black list of misleading websites. They found that compared with real news, the fake stories tended to be shorter and more repetitive with adverbs. Fake stories also had fewer outside quotes from sources, fewer technical words and fewer nouns.   I promise you will not find such dearth here at Brane Space.

Another indicator proposed by computer scientist Giovanni Luca Ciampaglia of Indiana University entails using a program to check how closely the subject and object f a statement are. Specifically, the fewer degrees of separation between a statement's subject ad object, and the more specific the intermediate words connecting them, the more likely the program is to label the statement as true.  To fix ideas consider the claim: "Obama is a Muslim" .  But the computer found that seven degrees of separation between the subject (Obama) and Muslim (e.g. Islamic) are just too many. Indeed, this nonsense was exploded  by PLOS One in fact checker back in 2015.  It was blown up even earlier (2009) by FactCheck.org in 2009, e.g.

Truth on the Cutting Room Floor, Dec. 4, 2009

A similar demolition was done on the Uranium One BS which has the Right's mindless parrots and screechers in a perpetual fit, e.g..


But it often isn't so easy especially when hate speech is embedded into fake news.  As pointed out in a piece yesterday (WSJ, p. A5) a post that is hateful to one group may not be hateful to another. Indeed, it may be perceived as "depicting uncomfortable truths".  This is inevitably what transpires when a large segment of the nation gets its news from FOX - which also comes with a good helping of hate, misinformation.

As another case showing how difficult it can be to get rid of a fake news nuisance, the article notes (ibid.):

"When Twitter executives struggled with whether to follow other Silicon Valley companies and remove conspiracy theorist Alex Jones from the platform in August, Twitter CEO Jack. Dorsey sought counsel from 'Ali Akbar' a conservative political activist".


Well, the very name of this dude  - a takeoff on the universal cry of jihadis everywhere ('Allahu Akbar!" including ISIS vermin) ought to have given Dorsey pause.   So no surprise Mr. "Akbar" advised Dorsey against kicking off Jones despite pressure from users and Twitter employees.  And let's be clear here that calling Jones a "conspiracy theorist"  confers way more gravitas than he merits.
Image result for alex jones
Alex Jones, conspiracy crackpot and hate monger.

As I've posted previously Jones is more a crackpot, pseudo conspiracy clown.  This is a person who not only lacks evidence for his claims, but who makes them in contradiction to all logic, mental coherence and existing documents. (Like Jones claiming the Sandy Hook massacre was a federal false flag operation and those 20 kids weren't really slain, just acting.)  

But without the ability to remove Jones and his irresponsible hate - crackpot conspiracy bollocks, users of the platform are basically on their own. Not only in filtering out the rot, but in actually discerning what is rot.   Am I overreacting in calling out Jones' content as hate? Judge for yourself based on this Jones outburst from 2016 in which  he fulminated on a crypto Jewish conspiracy afoot in the country:

I mean it’s not that Jews are bad; it’s just they are the head of the Jewish Mafia in the United States. They run Uber. They run the health care. They’re going to scam you. They’re going to hurt you.” 

Trump is roughly in the same balmy,  nutso camp as Jones with his "rigged election" conspiracy balderdash spouted during the 2016 campaign.  Like many loons he believes that merely repeating or tweeting unsupported claims makes them so. Thus his incessant references to his having "really" won the popular vote if there had been no  "voter fraud", i.e. by illegal immigrants.   

And let's note that Dotard's  own fake news has now reached  a pinnacle with his unsupported claims there is a "crisis" on the border which requires declaring a "national emergency".    He plans to visit that border near McAllen, TX tomorrow but be assured he will see no terrorists, or people rushing fences.  Never mind.  He will blurt any fantasy or brain fart  that emanates from a ruptured psychotic embolism in his amygdala.  He can't help himself. As I said before, in a normal universe the guy would be strapped down for ECT in a rubber room someplace, not sitting in the Oval Office barking bullshit or tweeting  twaddle.

What's really downright hilarious is how some, self-described Right "resistance"  backers of Trump are claiming the Left are full of "nutburgers" when it comes to twisting the truth. (Daniel Henninger WSJ column today, p. A13).  In fact a study by Princeton University researchers (Denver Post, today, p. 16A) showed:
Image may contain: text

The simplest reason for too many oldsters spouting or circulating fake news is the lack of digital literacy, according to one NYU researcher quoted in the piece.  The basis for  fake news spread by the ultra-conservatives is fairly simple too: they imbibe too much bullshit from Trump TV (aka FOX News), Breitbart, and Rush Limbaugh.

But the specter of the fake and its insinuation into every nook and crevice of the we, is even more insidious than the preceding examples, as Chris Hayes showed in his Dec. 29 show segment "Nothing is Real".   As Hayes stated, the Russian Facebook interventions in the 2016 election with reams of spurious stories about Hillary and Black Lives Matter, e.g.

http://brane-space.blogspot.com/2017/10/facebook-follies-yep-russkies-used-fake.html

Was just the "tip of the iceberg".  Guest Max Read in the short time (10 mins.) available touched base multiple times with his article in New York magazine about the extent of fake stuff on the internet, e.g.

http://nymag.com/intelligencer/2018/12/how-much-of-the-internet-is-fake.html

As he told Hayes: "One particular effort sponsored fake websites, and fake users - based on using bots, leading to nearly 40 percent of the web being fake."

Confirmed in the article wherein he wrote:

"Fake people with fake cookies and fake social-media accounts, fake-moving their fake cursors, fake-clicking on fake websites — the fraudsters had essentially created a simulacrum of the internet, where the only real things were the ads....

How much of the internet is fake? Studies generally suggest that, year after year, less than 60 percent of web traffic is human; some years, according to some researchers, a healthy majority of it is bot. For a period of time in 2013, The Times reported this year, a full half of YouTube traffic was “bots masquerading as people,” a portion so high that employees feared an inflection point after which YouTube’s systems for detecting fraudulent traffic would begin to regard bot traffic as real and human traffic as fake. They called this hypothetical event “the Inversion.”

The bots, in other words, taking control in this hypothetical inversion, wherein human traffic is taken as fake. Unbelievable? Not according to Read.  Much of this is being driven by defrauding ads which help pay for content, by making it appear as if there are actually human eyeballs on the ads when there aren't.. As Read put it to Chris Hayes:

 "In one of these cases the DOJ just indicted these guys with a rental service based in Dallas. They had built a whole raft of fake websites and created a bunch of fake people to go to the websites. They moved cursors like real people, they clicked on stuff like real people.  All of them bots. They even gave them social media accounts and had them surf the web to draw in cookies. So when they visited these fake websites it looked like you or me were visiting these sites.  Then they'd go to advertisers and say 'well hundreds of thousands of people have looked at your ad, now pay up based on your contract."

If Read's examination is true, and there's no reason to doubt it given the DOJ was involved, then it may mean the only way to clear out the fakes in the end is to have a truly free internet. NO more advertising just to see content, and no more capitalist yen to "monetize" the whole thing.  But I guess that's a non- starter given it's blatant "socialism".    A freebie socialism or human web traffic ultimately designated as fake: take your pick. 

But it's evident from Read's piece that as long as ads are tied to viewing web content many will try to "game the numbers".  Like legalizing drugs to stamp out or diminish the black market, it may be time to consider the extreme of no more ads, period.  At least then thinking human will have a better chance to decipher fake news using the methods already discussed - as well as good, old fashioned critical thinking!  

As for Brane Space, I ceased  including all ads (via  Google Adsense) some four months ago after discovering ads from Newsmax, etc. and other rightist sources  with snide attacks on Hillary, Dems, Obama. NO way in hell, as I told Janice, even if it means no more money for views. 

See also:



And:



by William Rivers Pitt | January 10, 2019 - 7:41am | permalink

No comments: