All posts by P. Takis Metaxas

Critical Analysis of News Propagation Online: TwitterTrails Classroom Activity

Uncategorized

We developed an educational activity for our Introduction to Media Arts and Sciences course “Computing for the Socio-Techno Web”, for students to think critically about the truthfulness of news propagated in social media. This activity utilizes TwitterTrails, a visual tool to analyze Twitter claims, events, and memes. This tool provides views such as a propagation graph of a story’s bursting activity, and co-ReTweeted network. Using a response and reflection form, students are guided through these different facets of a story.

We hope that other educators will further improve and use this activity with their own students.

Learning Goals

We defined the following learning goals for our in-class activity. In particular, we expect that following participation in the activity, students will be able to:

  1.  Understand the concepts of rumor spreading, including the extent and mechanisms in which stories propagate over social media.
  2.  Read and interpret visualizations that:
    Describe propagation of information over time
  3.  Conduct an evidence-based inquiry into the reliability of online information, employing a set of questions to examine who is spreading a story, and when and how was the story propagated
  4.  Identify indicators and characteristics that impact reliability including polarization, echo chamber, timing, and Twitters bots.

Activity materials

  • We presented the students with an 11-minute short demo on how to use TwitterTrails and we recorded the presentation in a video posted on YouTube.
  • Groups of 3-4 students selected randomly two of 12 stories and completed the Activity form.
  • Each student at the end completed a Reflection form on their experience investigating the stories.

Feel free to contact us, we would love to hear from you if you want to use our activity materials!

 

Evidence of pizzagate conspiracy theory on TwitterTrails

"fake news", Case Studies, conspiracy theory, Elections, Spammers

Many things have been written about the infamous #pizzagate conspiracy theory (“scandal” for those who believed in it) and, we are sure, much more will be written in the future. The fact that the outrageous and sick imagination of a few online trolls was able to persuade thousands of people that it was real, and motivate one of them to walk into the Comet Pizza with a loaded gun will be a matter of study of many Psychologists, Sociologists, and Political Scientists.  Given that in a few days a workshop will be held in Montreal on Digital Misinformation, we thought that this would be a good time to share some notes of a TwitterTrails story we did.

On December 2, 2016, we did a TwitterTrails investigation collecting Twitter data that contained the hashtag #pizzagate and we present here a few interesting observations about how it spread, who used it first and what were the shape of the community that engaged in the spreading until that time. Below are some of our findings. You can always explore the TwitterTrails story on your own.

Who used the hashtag #pizzagate first?

According to our data, the first mention of #pizzagate was at 8:34 AM GMT on Nov. 6, 2016, two days before the US elections. While the vast majority of people in the US were sleeping, the tweet was sent by a troll that has promoted tens of thousands of provocative lies to its 2 thousand followers. Most of the followers are certainly bots designed to infiltrate online groups willing to believe them — in this case Trump supporters.

If you want to know more about how trolls and spammers are successful in promoting lies, take a look at The Real “Fake News” post by Prof. Eni Mustafaraj.

 

Who made #pizzagate widely known?

The propagation graph below shows who were the main propagators of a rumor when its activity showed its first “burst”.

Clicking on the (partially covered) purple data point in the upper right corner we find that, surprisingly, the first tweet that had over 3000 retweets belongs to a pro-Erdogan Turkish journalist! According to The Daily Dot columnist Efe Sozeri, at that time, Turkey was outraged by a child abuse scandal and from controversial pending legislation on child marriage and governmental sources were trying to show that their scandal was minor compared to the US scandal.

But who informed the Turkish journalist about pizzagate? The propagation graph has some evidence that he was informed by a barrage of tweets that occur a few hours before his posting. The colorful column of data points just before his tweets are by a troll that sending dozens of tweets in Turkish. Here are a few of them as recorded in TwitterTrails:

 

A deafening echo chamber

The twitter exchange related to the pizzagate co-retweeted graph shows a dense echo chamber that is just verifying to its participants the validity of the conspiracy and allows no doubt to emerge:

This is the densest echo chamber we have observed on TwitterTrails. Among the 22,000 accounts posting about pizzagate, 4528 of them have risen to prominence being retweeted by at least two other accounts over a million times! Looking at the word cloud that characterizes the cyan group of 4474 participants, we see that the most common words in their profile are #maga, trump, truth [sic], love, god conservative.

How different is this graph from other graphs of political discourse? For comparison, we show what a typical co-retweeted network looks like when discussing political issues. Below is the graph related to the 2016 vice-presidential debate:

In this graph, you can see the two communities, their polarization, and the partial overlap as people read both sides but prefer one of them.

What else can we find?

These are just some of the insights that TwitterTrails can offer to a journalist or anyone who might want to study the propagation of a rumor. If you want to study it further, use TwitterTrails story of the hashtag #pizzagate and send us a comment!

 

 

The Real “Fake News”

"fake news", False Claims, Spammers

The following is a blog post that Eni Mustafaraj has recently published in The Spoke. We reproduce it here with permission.

Fake news has always been with us, starting with The Great Moon Hoax in 1835. What is different now is the existence of a mass medium, the Web, that allows anyone to financially benefit from it.

Etymologists typically track the change of a word’s meaning over decades, sometimes even over centuries. Currently, however, they find themselves observing a new president and his administration redefine words and phrases on a daily basis. Case in point: “fake news.” One would have to look hard to find an American who hasn’t heard this phrase in recent months. The president loves to apply it as a label to news organizations that he doesn’t agree with.

But right before its most recent incarnation, the phrase “fake news” had a different meaning. It referred to factually incorrect stories appearing on websites with names such as DenverGuardian.com or TrumpVision365.com that mushroomed in the weeks leading up to the 2016 U.S. Presidential Election. One such story—”FBI agent suspected in Hillary email leaks found dead in apparent murder-suicide”—was shared more than a half million times on Facebook, despite being entirely false. The website that published it, DenverGuardian.com, was operated by a man named Jestin Coler, who, when tracked down by persistent NPR reporters after the election, admitted to being a liberal who “enjoyed making a mess of the people that share the content”. He didn’t have any regrets.

Why did fake news flourish before the election? There are too many hypotheses to settle on a single explanation. Economists would explain it in terms of supply and demand. Initially, there were only a few such websites, but their creators noticed that sharing fake news stories on Facebook generated considerable pageviews (the number of visits on the page) for them. Their obvious conclusion: there was a demand for sensational political news from a sizeable portion of the web-browsing public. Because pageviews can be monetized by running Google ads alongside the fake stories, the response was swift: an industry of fake news websites grew quickly to supply fake content and feed the public’s demand. The creators of this content were scattered all over the world. As BuzzFeed reported, a cluster of more than 100 fake news websites was run by individuals in the remote town of Ceres, in the Former Yugoslav Republic of Macedonia.

How did the people in Macedonia manage to spread their fake stories on Facebook and earn thousands of dollars in the process? In addition to creating a cluster of fake news websites, they also created fake Facebook accounts that looked like real people and then had these accounts subscribe to real Facebook groups, such as “Hispanics for Trump” or “San Diego Berniecrats”, where conversations about the election were taking place. Every time the fake news websites published a new story, the fictitious accounts would share them in the Facebook groups they had joined. The real people in the groups would then start spreading the fake news article among their Facebook followers, successfully completing the misinformation cycle. These misinformation-spreading techniques were already known to researchers, but not to the public at large. My colleague Takis Metaxas and I discovered and documented one such techniqueused on Twitter all the way back in the 2010 Massachusetts Senate election between Martha Coakley and Scott Brown.

There is an important takeaway here for all of us: fake news doesn’t become dangerous because it’s created or because it is published; it becomes dangerous when members of the public decide that the news is worth spreading. The most ingenious part of spreading fake news is the step of “infiltrating” groups of people who are most susceptible to the story and will fall for it.  As explained inthis news article, the Macedonians tried different political Facebook groups, before finally settling on pro-Trump supporters.

Once “fake news” entered Facebook’s ecosystem, it was easy for people who agreed with the story and were compelled by the clickbait nature of the headlines to spread it organically. Often these stories made it to the Facebook’s Trending News list. The top 20 fake news stories about the election received approximately 8.7 million views on Facebook, 1.4 million more views than the top 20 real news stories from 19 of the major news websites (CNN, New York Times, etc.), as an analysis by BuzzFeed News demonstrated. Facebook initially resisted the accusation that its platform had enabled fake news to flourish. However, after weeks of intense pressure from media and its user base, it introduced a series of changes to its interface to mitigate the impact of fake news. These include involving third-party fact-checkers to assign a “Disputed” label to posts with untrue claims, suppressing posts with such a label (making them less visible and less spreadable) and allowing users to flag stories as fake news.

It’s too early to assess the effect these changes will have on the sharing behavior of Facebook users. In the meantime, the fake news industry is targeting a new audience: the liberal voters. In March, the fake quote “It’s better for our budget if a cancer patient dies more quickly,” attributed to Tom Price, the Secretary of Health and Human Services, appeared on a website titled US Political News, operated by an individual in Kosovo. The story was shared over 80,000 times on Facebook.

Fake news has always been with us, starting with The Great Moon Hoax in 1835. What is different now is the existence of a mass medium, the Web, that allows anyone to monetize content through advertising. Since the cost of producing fake news is negligible, and the monetary rewards substantial, fake news is likely to persist. The journey that fake news takes only begins with its publication. We, the reading public who share these stories, triggered by headlines engineered to make us feel outraged or elated, are the ones who take the news on its journey. Let us all learn to resist such sharing impulses.

False and True rumors look very differently on TwitterTrails

About TwitterTrails, Case Studies, False Claims, True Claims

News of Turkish airplane shooting down a Russian one over the Turkish-Syrian border has dominated the news and the social media lately. We investigated the rumor within hours after it appeared (24 Nov. 2015) and you can see the ressults of the analysis here: http://twittertrails.wellesley.edu/~trails/stories/investigate.php?id=462776628

This was not the first time a rumor of this kind emerged. About a month and a half ago (10 Oct. 2015) an identical rumor had emerged. We had investigated that rumor too and you can see the results of our anaysis here: http://twittertrails.wellesley.edu/~trails/stories/investigate.php?id=134661966

Russian jet downing rumors

As you can see, based on the crowd’s reaction to the rumors, TwitterTrails was able to determine that the October rumor was false while the November one was true. The false rumor did not spread much and had a lot of skeptical tweets questioning its validity. On the other hand, the true rumor spread much higher and in terms of skepticism was undisputed.

Our understanding of the way the “wisdom of the crowd” works is that, when unbiased, emotionally cool observers see a rumor that seems suspicious, they usually react in one of two ways: They either do not retweet it, reducing its spread, or they may respond questioning the validity of the rumor, resulting in higher skepticism.

Continue reading False and True rumors look very differently on TwitterTrails →

Following #GRexit on Twitter Trails

#GRexit, Events and Memes

TwitterTrails is a system that can easily follow the spread of any set of words on Twitter, including hashtags. Today we followed the hashtag #GRexit, associated with the issue of whether Greece will eventually exit the Eurozone or not. For those following the Greek drama to secure new loans and remain in the Eurozone, it appears that mentioning GRexit has increased in the media in the past month. Discussions about GRexit go back to 2010 but there was never before considered as likely an event  to happen as it is now.

Increased mentioning of #GRexit

We  collected tweets containing the hashtag #GREXIT on TwitterTrails.com today (warning: this is a large data set and it takes a while to load). Below are some observations from this first collection:

GRexit timeline

Looking at the time series of the collection, there is, indeed, an increase in the appearance of #GRexit since the Greek elections in January, 25 2015, but the first tweet in our data set goes back almost 3 year: Interestingly, people remember the tweet sent by the then-Economist/Blogger Yianis Varoufakis in May 2012. In this tweet, addressing two reporters,  Mr. Varoufakis gives a psychological definition of the term (my translation from Greek):

“Grexit: The unsatisfiable desire of Germany (which, if satisfied, euro collapses). That simple.”

Not sure what Mr. Varoufakis meant by this tweet — let me know if you do. Through his blog and interviews, Mr. Varoufakis is known for writing and saying a lot of things over time.  I recall listening to him giving one of his many interviews to NPR in July, 2012, telling the reporter in no uncertain terms that the euro will not exist by the end of August, 2012. Anyway, whatever he meant back then, he is now Finance Minister of Greece and he really has to avoid the possibility of GRexit.

Current visibility of #GRexit

Next, let’s take a look at the tweet Propagation Graph showing what made #GRexit popular recently.

GRexit propagation graph

The tweet propagation graph is showing which tweets made the hashtag popular. The highest point in the propagation graph comes from BBC News (World) on 17 Feb, 2015 (the large purple circle) discussing the implications of a GRexit: “What would happen if Greece quits the euro? #grexit“. The attached picture in the tweet shows  drachma coins and bills, the expected new currency if Greece were to exit the Euro.

BBC highest tweet

 

How is the audience following #GRexit

Next, we look at the co-retweeted network (below). Recall that the co-Retweeted network is showing the “main players” among the accounts tweeting about #GRexit, according to the audience. Not surprising, this network is mainly composed of news agencies in European countries. The shape of the network is important as it reveals the degree of polarization of the audience following GRexit.

The more often two accounts are being co-retweeted by others, the closer they appear.
The further they appear, the less they are co-retweeted.

GRexit co-retweeted networkThe group shown at the top of the network (colored yellow) is composed mainly of German news agencies.
At the other end, the bottom group (colored pink) is composed mainly of pro-Greek government accounts, and near  them (colored light green) Greek news agencies. They are as far apart as they can be.
In the middle groups (colored orange, blue and light green), are news agencies from other European countries.

One way to read the co-retweeted network is that the audience in Greece and Germany are exposed to different news. This makes some sense since it is unlikely that news agencies tweet mainly in the language of their audience. But it may also mean that the audiences are exposed to different version of events and are lost in translation. There were at least two misunderstandings that were blamed to inaccurate translation.

We will continue to monitor GRexit on a bi-weekly basis. If you are interested in knowing what happens to the popularity and visibility of the hashtag, stay tuned!

 

 

 

According to the crowd, Putin’s motorcade shape was a hoax

Case Studies, False Claims

Twittertrails is studying the propagation of rumors on Twitter. It will give you evidence about how the Twitter audience reacts to a rumor and whether the audience believes the rumor is a hoax or not. Its method of measuring the skepticism of the audience is a little more sophisticated than counting retweets, however. In fact, the rumor may get a huge head start, almost 11,000 retweets, but if they all came from the same source, they will not count as much on the skepticism level.

Let us demonstrate the skepticism level with an example. The twittertrails.com user asks the system to retrieve all tweets containing the terms “putin motorcade”, and the rest are computed automatically in a few minutes.

The “Putin Motorcade Shape” Story

A few days ago, a funny rumor appeared, that Putin’s motorcade was shaped as, well, you’ll see in the accompanying picture:

putin-propagation

The originating tweet who broke the story received almost 11,000 retweets! One of the reasons for that amazing count that any marketer would love, was that the account owner is a member of several groups with different political, ethnic and financial identities. Looking into the timeline graph, however, one can see that he was not the first to write about it. The first relevant tweet came several days before his post, by an Ukranian Euromaidan account.

Continue reading According to the crowd, Putin’s motorcade shape was a hoax →

False rumors do not spread like True ones

About TwitterTrails

On Twitter, claims that receive higher skepticism and lower spread scores are more likely to be false.
On the other hand, claims that receive lower skepticism and higher spread scores are more likely to be true.

The above is a conjecture we wrote in a recent paper entitled Investigating Rumor Propagation with TwitterTrails (currently under review). Feel free to take a look if you want to know more details about our system, but we will write here some of its highlights.

As you may know if you have read our Twitter Trails Blog before, we are developing a Web service that, starting from a tweet or a set of keywords related to a story spreading on Twitter (or a hashtag), it will investigate it and answer automatically some of the basic questions regarding the story. If you are not familiar, you may want to take a look at some of the posts. Or, it can wait until you read this one.

Recently we deployed twittertrails.com a site containing the growing collection of stories and rumors that we investigate. Its front end looks like this:

condensed_view_v2

 

This is the “condensed view” which allocates one line per story, 20 stories per page. There are over 120 stories collected at this point. Clicking on a title brings you the investigation page with lots of details and visualizations about its spread, its originator, how it burst, who supports it and who refutes it.

Continue reading False rumors do not spread like True ones →