Two computer science students create tool to detect “bots” on Twitter

Two computer science students created a Google Chrome extension that when clicked tells you if a Twitter user appears to be a bot or not.

They claim it has 93.5% accuracy[1] (but see the footnote for a hint at some of the problems in how they came to that conclusion). It uses “machine learning” technology to attempt to identify Twitter accounts that may be automated “propaganda” accounts. Per the article, their classifier was trained using Tweets identified as left or right leaning – and those which they could not categorize as left or right must be bots. Or something. Regardless, that implies political views play a role in classification as a bot. Would a bot tweeting about cats be identified? Would a propaganda bot promoting backyard gardening be identified?

The results could be manipulated by users. When the bot check reports its results to you, you can optionally agree or disagree – and that information gets fed back to the classifier. A sufficient number of users could likely re-train the classifier to intentionally classify real people as bots, and bots as real people.

Source: The College Kids Doing What Twitter Won’t | WIRED

I am not convinced that software tools can classify propaganda bots with sufficient accuracy to be useful over the long term. There will be an arms race to create better bots that appear more natural. I fear that such tools may be used to stifle speech by incorrectly – or deliberately – classifying legitimate speech as “bot” generated to have that speech throttled down or banned.

Note also that Twitter – and Facebook – profit by having emotionally engaged users reading, liking, sharing and following more people. It is not yet in their financial interest to be aggressive about shutting down bots.


How good is 93.5% accuracy? Let’s consider a different example to understand this: the use of drug search dogs in schools to locate drugs in lockers.

Let’s say the dog has a 98% accuracy in finding drugs in a locker and a 2% false positive rate. Further, let’s assume there are 2,000 lockers in the school.

Let’s assume 1% of the students actually have drugs in their locker.

1% of 2000 students means 20 students actually have drugs in their locker. (And with the 2% false rate there is a chance that 1 of these actual students will be missed.)

In using the dog, the police will identify that 2% (the false positive rate) of the lockers incorrectly or 40 lockers will be suspected of having drugs in a school where only 20 lockers have drugs.

In other words, twice as many students will be falsely accused of having drugs as students who actually have drugs.

When doing broad classification searches, even a 98% accuracy rate is problematic as it may produce more false negatives than true positives, which is not what you would intuitively guess when you hear “98% accuracy” or in this Twitter bot analysis, 93.5% accuracy.

Further, in determining their 93.5% figure, while their approach is admirable and possibly the best that can be done, they compared verified Twitter user tweets to suspected “bots” from unverified accounts. Most Twitter accounts are unverified and they are only hypothesizing that an account is a bot when producing this metric. (FYI I think they have done an excellent job with their work, the best that can be done, and am impressed with their work. My comments should in no way be interpreted as negative comments towards these two students. For the record, I have a degree in computer science and an M.S. in software engineering and have familiarity – but not expertise – in machine learning, classifiers and pattern matching systems.)

Indeed, as the article points out, hundreds of people have already complained to another bot checker about being falsely classified as a bot. The Wired reporter attempted to contact the account holders of a small sample of accounts identified as bots and quickly found accounts that appeared to be run by real people.

Side note: the linked article in Wired is excellent journalism, something I certainly do not see enough of! Glad to see this article!

How “Bot Armies” get Twitter hashtags trending

Of interest, a bot army is said to have “taken to Twitter” to influence Twitter social media posts. Bots generate enough Tweets that eventually get shared and then turn into actual hashtag memes passed along by real people. In this way, propaganda bots can initiate and control messaging on social media.

This is also known as “computational propaganda”. In the old days, propaganda usually required a printing press or a broadcast license. Social media made it possible for everyone to be a propagandist. Computational propaganda creates fully automated propaganda dissemination.

Source: Pro-Gun Russian Bots Flood Twitter After Parkland Shooting | WIRED


The story behind the fake DPRK News Service Twitter account

The DPRK News Service is one of the funniest Twitter accounts in existence as it hilariously makes fun of North Korea and a host of world leaders and events.

It’s run by 2 guys in the U.S. who launched it to share jokes between friends.

Not long after launching it, the U.S. media repeatedly cites them as an official North Korean news service.

“It feels like all online media outlets have prioritized being First rather than Right, like it’s some sort of Scoop Race. As mentioned before, a cursory Google search will reveal us instantly, or even a cursory look at other tweets should reveal some level of incredulity. I imagine when some poor journalist is banging out some listicle or whatever on the 10 Ways North Korea is a Horrific Nightmare of Human Rights Abuses, they’re not going to really question things if some twitter account with a sizeable follower count blurts out something that fits within the context of the article.”

Source: The Men Behind the Infamous Fake North Korean Twitter Account

Yesterday, reporter Michelle Kosinski, who has a history of making up news stories, re-tweeted a DPRK News Service tweet.

She is famous for her live report from a dangerous flood, while sitting in a canoe – in what embarrassingly turns out to be about six inches of water. A clear example of fake news reporting.

Watch this short news clip to see a professional journalist honestly reporting the news and protecting Democracy itself… not. When you see this, keep in mind how much “news” is just story telling – and wrap your mind around the role propagandists play in shaping the stories presented as “news”. Journalists should serve as propaganda firewalls protecting viewers and readers from the onslaught but many play along and act as propagandists themselves, promoting their own agendas.

After reading DPRK News Service hilarious parodies, try to read the news headlines without laughing! It’s hard!

Newsweek fires all top staff


Newsweek on Monday fired all of its top staff amid turmoil that has upended the newsroom .In a company meeting, several editors announced that the outlet had fired Editor in Chief Bob Roe, Executive Editor Ken Li and reporters Celeste Katz, Josh Saul, and International Business Times editor Josh Keefe.

Source: Newsweek Guts Its Top Edit Staff Amid Legal Turmoil

Newsweek has a reputation as fake news publisher.

We use the definition of fake news as the use of exaggerated, emotionally laden, often false, click-bait reports as means of generating Likes and Shares for selling eyeballs to advertisers.

Newsweek is discovering that pretending to be a legitimate news service is not the best way to run a news business.

Last time Newsweek went up for sale, it sold for one dollar. What do you think its market value is today?

When fake news is not good enough, buy fake readers to sell to advertisers

This blog has long suggested that Newsweek appears to be based on the fake news publishing model. Now we learn their parent company has fake readers:

Newsweek Media Group, the parent company of IBT, acknowledged it purchases audiences from ad networks that sell pop-up and pop-under traffic

Source: The Publisher of Newsweek And The International Business Times Has Been Buying Traffic And Engaging In Ad Fraud

The head editor at Newsweek is currently suspended over allegations of sexual harassment. The IRS is said to have filed millions of dollars in tax liens against the publisher.

Fake news publishers often intermix real news with exaggerated, emotion-laden click-bait articles designed for online sharing on social media. The goal is to drive eyeballs to advertisers. And when that is insufficient to create ad traffic, they can always buy fake readers.

To avoid fake news on Facebook, just avoid all news on Facebook?

According to the Pew Research Center, Americans are between two and three times more likely to trust the local news (82 percent) and the national news (76 percent) than they are to trust anything they see on social media (34 percent).

Source: How to avoid fake news: Stop getting your news from Facebook, that’s how

Corollary: To avoid spreading fake news, only share content that you create yourself or that which is created by people you personally know.

In spite of the propaganda spin, it really does take 60 votes, not 51, to approve the Federal budget

Online, social media-based, for profit, fake news publisher Occupy Democrats sent out another false meme for viral distribution on social media. This item was shared into my Facebook news feed.

The Republicans have a 52 seat majority in the Senate, with one absent for medical treatment, giving them a 51 seat majority.

The problem with this propaganda is that due to Senate rules and how the political parties work in real life, it takes 60 votes, not 51.

CSPAN understood this with this graphic after the vote to continue funding the government failed on January 19, 2018 – note the reference to “60 votes needed to limit debate”.

CNN explained this in a news report just prior to the vote:

“Sixty votes were needed to advance the bill. Republicans only control 51 seats, so GOP leaders needed Democratic votes to cross that threshold. As of 11 p.m. ET, the vote was still technically ongoing but enough senators had voted against the plan to prevent it from advancing. “

CNBC noted that it takes 60 votes to pass.

The day after, the Associated Press described the situation with this headline:

Which is the opposite of the propaganda poster, above.

In 2012, Obama’s Chief of Staff said that it took 60 votes. Politifact took issue with the exact wording Obama’s Chief of Staff used and said a budget resolution can technically pass with 51 votes, but agreed that in reality, it takes 60 votes to move past the resolution phase.

Because of wide spread propaganda from fake news sites, social media on Twitter and Facebook are filled with posts saying 51 votes or 60 votes are sufficient, depending on the partisan bias of the poster. The battle crossed onto social media where it basically slid down hill from there.

Facebook and Google announce they are launching their own anti-propaganda content

Anti-propaganda is just propaganda in front of a mirror.

Facebook, Google and Twitter say they’re creating more anti-terror propaganda to combat violent messages at their source.

Source: Facebook, Google tell Congress how they’re fighting extremist content

Google’s Youtube analyzes your past searches and viewing patterns and “determines” if the viewer “may be headed toward extremism”. If so, Youtube presents ads to “subtly contradict” others’ propaganda messages.

Youtube is removing comments from videos that contain hateful content, as determined by Youtube’s “machine learning” algorithms, automatically. Not mentioned here, but Youtube has also gone to “de-monetization” of videos containing video content that Youtube does not like. What Youtube does not like is hard to determine – I watched a video by a photographer showing that Youtube is “de-monetizing” his photography “how to” videos, for no apparent reason.

A related Youtube change effective immediately, is that Youtube channels are prohibited from running advertising if they have less than 1,000 subscribers and at least 4,000 hours of content viewed during the preceding year.

Facebook is supporting “counterspeech efforts” using automated image and language analysis to identify content. Facebook and Google both say they are using propaganda methods to counter program those they think are receiving propaganda from “hateful” or “violent” persons or groups.

Twitter is using “Tweet throttling”, a form of shadow banning, and down marking trending topics that Twitter decides should not be trending.

Facebook to alter its news feed sharing algorithms

Facebook is said to be rolling out changes to how content is selected for each of us to see in our news feed. FB will show us content that their algorithms believe is user generated (versus that from publishers), and which has been shared, liked or commented upon. Other posts will apparently see lower priority and less visibility, meaning, they sort of disappear.

Facebook says this should result in seeing more personal posts from actual FB friends. FB is trying to discourage “passive” reading of posts and wants to push people to interact more. I suspect they are pushing towards interaction because passive reading provides them no data on your interests. By leading you to click Like, Share or Comment, they can detect your interests which they use to refine their dossier on each of us, to improve the marketing of products and services to us.

Source: ‘We’re losing hope’: Facebook tells publishers big change is coming to News Feed – Digiday

The notionally accurate NY Times

This tweet is not the way to respond to accusations of “fake news”.

Maggie Haberman is the White House reporter for the NY Times. Perhaps she does not know what the word “notional” means – or, she is advocating fake news is fine as long as meets (her) imaginary world view.


  1. pertaining to or expressing a notion or idea.
  2. of the nature of a notion or idea: a notional response to the question.
  3. abstract, theoretical, or speculative, as reflective thought.
  4. not real or actual; ideal or imaginary: to create a notional world for oneself.
  5. given to or full of foolish or fanciful ideas or moods.
In other words, Haberman is saying that though it is false, it meets our imaginary idea of what we want it to be, therefore, it is true.
It is perplexing why the media does this to itself. I do not think Orwell intended his novel “1984” to be a “how-to” guide for the NY Times.