A few days ago the Center for Research on Computation and Society organized a workshop with the provocative title “Six Reasons Fake News is the End of the World as we Know It“. I call it provocative because, whether “fake news” is a new thing or not, has been discussed a lot lately. Not all of us agree on what it is, or how novel it is. Some point out that it is as old as newspapers, others see it as something that mainly appeared last year. Yet others doubt that it is even a phenomenon worth discussing and that, instead of fake news, we should talk instead about specific categories such as false news, misinformation, disinformation, and propaganda.

Accepting the challenge, I gave a talk with an equally provocative, I would like to believe, title:  “Artificial Intelligence, your brain, and other things you cannot trust about politics“. You can follow my talk in the video below, but let me give you a list of the “things” that I discussed in the talk:

I hope you find it interesting and do your own thinking about what we can trust when it comes to politics. Importantly, we need to figure out how to solve the problems of online misinformation and propaganda that seem to be all around us these days.

Or, to learn how to live with them, which is what I think will happen.

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 technique used 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 in this 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.

I recently had the uninvited opportunity to defend my domain name, metaxa.net, and I am writing this post because it may be useful to others who find themselves in such a position. The good news: You do not need to be a lawyer to do it yourself and the arbitration system works reasonably well. The bad news: You need to do a bit of reading and writing.

I registered and own metaxa.net since 1999. Back then it was essential to own a domain name since, at that time, there were no services to upload and share your photos, no easy ways to have email addresses for you or your family members, and the only clouds around were still up in the sky unable to store any files you may need while away from your office.

As the Web services expanded to cover everything under the sun, the above uses became secondary, and I started utilizing the domain name for other reasons. In 2004 I had started a line of research to discover how Web Spammers succeeded in gaming search engines and place their bogus postings in the top-10 page of relevant search results. Back then it was thought that the “PageRank” algorithm was like “42”: the answer to everything (related to the Web). We now know that search engines can be gamed, and that they spend considerable resources to avoid Web Spam.

[Side note: My research led me to discover the reasons that search engines can be gamed —  pretty much for the same reasons we, humans, can be fooled. Web Spammers were using techniques very similar to the propagandistic techniques that politicians, advertisers and financial criminals are using to persuade us to vote for something, buy something or invest in something. I presented my work initially at AIRWeb 2005 (Web Spam, Progaganda and Trust) but if you are interested to read more you should check the journal version (Web Spam, Social Propaganda and the Evolution of Search Engine Rankings).]

Anyway, to implement a technique of discovering Web Spam sites I needed a method to evaluate similarity between Web site contents. It helped to have a Web site containing text that is unbiased towards a particular theme or product. Since there was nothing available online, I uploaded onto my own site a large collection of Associated Press news that were used by the TREC community. If you visited the top directory of metaxa.net you would be surprised to find huge files containing old news. But it was very unlikely that you would visit it. I did not include metaxa.net in any search engine listing, so it would never appear in your search results. Though I never planned to use metaxa.net as text repository, over the years it proved very useful. Many of my students used it to run their research projects, and after graduation some of them ended up working at Microsoft Bing and Google Search, fighting Web Spam on a regular basis.

I continued to own the domain name paying the fees on time, so it was a complete surprise when I received a letter in early April about a dispute filed with the World Intellectual Property Organization (WIPO) by Remy-Cointreau Luxemburg, the well known liqueur producing company. They wanted to take over my domain name! Remy-Cointreau had bought a well-known Greek Spirits company, METAXA, and over the years they had started buying every domain name that contained the string “metaxa”  (including “metaxaswineestate”!) Given my research, I knew exactly why they were doing that: They wanted to “persuade” the search engines that any search of the term “metaxa” should lead to their official site only! They knew how to fool PageRank and were doing it legally too. Now they wanted to fool WIPO’s arbitrators. In fact, as they claimed in the Complaint they served me,

“Indeed, the term METAXA® is only known in relation to the Complainant. It has no meaning whatsoever in English or in any other language. A Google search on the term METAXA® displays several results, all of them being related to the Complainant“

Wow. Three lies in three sentences. Anyone who knows just a bit of modern Greek history or checks Wikipedia knows that the name Metaxas is not that rare. (In Greek, the female version of a name, or a reference to the family name itself, does not include the final “s”.) One can Google translate metaxa to see that it means “raw silk” and “silk trader” in Greek, depending on the intonation. And the first page of Google search results does not constitute a proof of unique association. (Yet, even their own submitted screenshot included other METAXA references!)  And these were not the only “inaccurate claims” or logical fallacies in the Complaint. There were about a dozen of them. You can see a more detailed list (though not exhaustive) in my Response to their Complaint.

I was doubly stunned. Using such claims were ridiculing the WIPO, the body that they were asking to support them. How can they be so arrogant insulting WIPO’s arbitrators’ intelligence? Wouldn’t they expect that the Responder would point out their lies?

Probably not. My guess is that they expected that nobody would respond to their Complaint and they would win an uncontested case. You see, when a couple years ago we changed the domain record hosting company, the WHOIS information was not updated correctly, and it showed that the administrator was to be reached at … nocontactsfound@secureserver.net. Seems reasonable to expect that whoever owned the domain would not be reached, and so they could snatch it without contest. It would cost them a few thousand dollars, but for a company with deep pockets, that would not be a problem.

Unfortunately for them, due to a billing inquiry, I did get informed. Legalese never being part of my tongues, I turned to the wonderful Berkman community for advice. And the advice poured in immediately. Several Berkmanites, and primarily Prof. Jacques de Werra of the University of Geneva and Faculty Associate at the Berkman Center this academic year, suggested literature, gave me references to other relevant cases, recommended experts in the field, offered advice on my options. They even pointed out to other unfair activities that the company was involved in the past. In particular, a song written for the company’s ad campaign was stolen from Berkmanite musician Erin McKeown – you can read all about it at TechDirt Case study A Perspective On The Complexities Of Copyright And Creativity From A Victim Of Infringement.

Onto the technical part. It turns out that one can defeat a Complaint by convincing the arbitrators that any of these elements are not present:

1. your domain name is identical or confusingly similar to a trademark or service mark in which the complainant has rights;
2. you have no rights or legitimate interests in respect of the domain name; and
3. your domain name has been registered and is being used in bad faith.

The detailed policy and rules (called a UDRP, Uniform Domain-Name Dispute-Resolution Policy) are listed here:
http://www.icann.org/en/help/dndr/udrp

In my case, the first element could not be countered: There were good reasons why my domain name was identical to theirs. But I could convince the arbitrators that the Complaint was wrong for both the second and the third element. That I have legitimate rights, was straight forward: I had to just point out the lies within their claims. But you need to provide a complete counter argument. It is not enough to point out a lie and expect that the arbitrators will go looking for its validity. You have to provide it yourself. With screenshots, excerpts, clear arguments. And you better be exhaustive in your arguments because you may only get one shot. For example, even though I was quite sure that I could nullify the second element, I should better also nullify the third one, just in case. For the third element, I had to go digging for references and receipts showing that I was always the owner of the account and had used it in good faith. Showing “good faith” was important, as this is a main reason why UDRPs exist: to curb the efforts of cyber-squatters who buy domain names only to sell them to the higher bidder of competing companies.

While I was at it, I wanted to point out that the Complaint itself was not filed in good faith. Remy-Cointreau really did not really have a case. Only through the dozen of lies in the filing they could put a case together. I would love to get the arbitrators acknowledge the Company’s filing with bad faith. There are no penalties associated with such an acknowledgement, but future arbitrators may take it into account in the future.

At the end the arbitrators denied the Remy Cointreau – Metaxa Complaint, stopping at the fact that the second element was not proven by the company. In addition to the company losing a case and a few thousand dollars, they lost the opportunity of persuading search engines about the unique association of their trademark. Now that this is an officially recorded WIPO UDRP case, it may help reducing the number of future frivolous Complaints.

PS. I recently found a good guide on How to Choose the Right Domain Name, I hope you will find it useful.

Dear Readers,

Below are my annotated notes from a talk I gave at Berkman’s Truthiness in Digital Media Symposium a few weeks ago. I introduced the concept of Social Theorems, as a way of formulating the findings of the research that is happening the last few years in the study of Social Media. It is my impression that, while we publish a lot of papers, write a lot of blogs and the journalists report often on this work, we have troubles communicating clearly our findings. I believe that we need both to clarify our findings (thus the Social Theorems), and to repeat experiments so that we know we have enough evidence on what we really find. I am working on a longer version of this blog and your feedback is welcome!

Best,
P. Takis Metaxas

With the development of the Social Web and the availability of data that are produced by humans, Scientists and Mathematicians have gotten an interest in studying issues traditionally interesting mainly to Social Scientists.

What we have also discovered is that Society is very different than Nature.

What do I mean by that? Natural phenomena are amenable to understanding using the scientific method and mathematical tools because they can be reproduced consistently every time. In the so-called STEM disciplines, we discover natural laws and mathematical theorems and keep building on our understanding of Nature. We can create hypotheses, design experiments and study their results, with the expectation that, when we repeat the experiments, the results will be substantially the same.

But when it comes to Social phenomena, we are far less clear about what tools and methods to use. We certainly use the ones we have used in Science, but they do not seem to produce the same concrete understanding that we enjoy with Nature. Humans may not always behave in the same, predictable ways and thus our experiments may not be easily reproducible.

What have we learned so far about Social phenomena from studying the data we collect in the Social Web? Below are three Social Theorems I have encountered in the research areas I am studying. I call them “Social Theorems” because, unlike mathematical Theorems, they are not expected to apply consistently in every situation; they apply most of the time and when enough attention has been paid by enough people. Proving Social Theorems involves providing enough evidence of their validity, along with description of their exceptions (situations that they do not apply). It is also important ti have a theory, an explanation, of why they are true. Disproving them involves showing that a significant number of counter examples exists. It is not enough to have a single counter example to disprove a social theorem, as people are able to create one just for fun. One has to show that at least a significant minority of all cases related to a Social Theorem are counter-examples.

SoThm 1. Re-tweets (unedited) about political issues indicate agreement, reveal communities of likely minded people.

SoThm 2. Given enough time and people’s attention, lies have short questioned lives.

SoThm 3. People with open minds and critical thinking abilities are better at figuring out truth than those without. (Technology can help in the process.)

So, what evidence do we have so far about the validity of these Social Theorems? Since this is simply a blog, I will try to outline the evidence with a couple of examples. I am currently working on a longer version of this blog, and your feedback is greatly appreciated.

Evidence for SoThm1.

There are a couple papers that present evidence that “Re-tweets (unedited) about political issues indicate agreement, reveal communities of likely minded people.” The first is the From Obscurity to Prominence in Minutes: Political Speech and Real-Time Search paper that I co-authored with Eni Mustafaraj and presented at the WebScience 2010 conference. When we looked at the most active 200 Twitter users who were tweeting about the 2010 MA Special Senatorial election (those who sent at least 100 tweets in the week before the elections), we found that their re-tweets were revealing their political affiliations. First, we completely characterized them into liberals and conservatives based on their profiles and their tweets. Then we looked at how they were retweeting. In fact, 99% of the conservatives were only retweeting other conservatives’ messages and 96% of liberals those of other liberals’.

Then we looked at the retweeting patterns of the 1000 most active accounts (those sent at least 30 tweets in the week before the elections) and we discovered the graph below:

As you may have guessed, the liberals and conservatives are re-tweeting mostly the messages of their own folk. In addition, it makes sense: The act of re-tweeting has the effect of spreading a message to your own followers. If a liberal or conservative re-tweets (=repeats a message without modification), he/she wants this message to spread. In a politically charged climate, e.g., before some important elections, he/she will not be willing to spread a message that he disagrees with.

The second evidence comes from the paper “Political Polarization on Twitter” by Conover et. al. presented at the 2011 ICWSM conference. The retweeting pattern, shown below, indicates also a highly polarized environment.

In both cases, the pattern of user behavior is not applying 100% of the time, but it does apply most of the time. That is what makes this a Social Theorem.

Evidence for SoThm2.

The “Given enough time and people’s attention, lies have short questioned lives” Social Theorem describes a more interesting phenomenon because people tend to worry that lies somehow are much more powerful than truths. This worry stems mostly from our wish that no lie ever wins out, though we each know several lies that have survived. (For example, one could claim that there are several major religions in existence today that are propagating major lies.)

In our networked world, things are better, the evidence indicates. The next table comes from the “Twitter Under Crisis: Can we trust what we RT?” paper by Mendoza et. al., presented at the SOMA2010 Meeting. The authors examine some of the false and true rumors circulated after the Chilean earthquake in 2010. What they found is that rumors about confirmed truths had very few “denies” and were not questioned much during their propagation. On the other hand, those about confirmed false rumors were both questioned a lot and were denied much more often (see the last two columns enclosed in red rectangles). Why does this make sense? Large crowds are not easily fooled as the research on crowd sourcing has indicated.

Again, these findings do not claim that no lies will ever propagate, but that they will be confronted, questioned, and denied by others as they propagate. By comparison, truths will have a very different experience in their propagation.

The next evidence comes from the London Riots in August 2011. At the time, members of the UK government accused Twitter of spreading rumors and suggested it should be restricted in crises like these. The team that collected and studied the rumor propagation on Twitter found that this was not the case: False rumors were again, short-lived and often questioned during the riots. In a great interactive tool, the Guardian shows in detail the propagation of 7 such false rumors. I am reproducing below an image of one of them, the interested reader should take a closer look at the Guardian link.

During the Truthiness symposium, another case was presented, one that supposedly shows the flip side of this social theorem: That “misinformation has longer life, further spread on Twitter than accompanying corrections”. I copy the graph that supposedly shows that, for reference.

Does this mean that the Social Theorem is wrong? Recall that a Social Theorem cannot be refuted by a single counter-example, but by demonstrating that, at least a significant minority of counter examples, exists.

Further, the above example may not be as bad as it looks initially. First, note that the graph shows that the false spreading had a short life, it did not last more than a few hours. Moreover, note that the false rumor’s spreading was curbed as soon as the correction came out (see the red vertical line just before 7:30PM). This indicates that the correction probably had a significant effect in curbing the false information, as it might have continue to spread at the same rate as it did before.

Evidence for SoThm3.

I must admit that “People with open minds and critical thinking abilities are better at figuring out truth than those without” is a Social Theorem that I would like to be correct, I believe it to be correct, but I am not sure on how exactly to measure it. It makes sense: After all our educational systems since the Enlightenment is based on it. But how exactly do you created controlled experiments to prove or disprove it?

Here, Dear Reader, I ask for your suggestions.

Dear Readers,

The following is a blog that I wrote recently for a conference on “Truthiness in Digital Media” that is organized by the Berkman Center in March. It summarizes some of the research findings that have shaped my approach to the serious challenges that misinformation propagation poses in Cyberspace.

Do you have examples of misinformation or propaganda that you have seen on the Web or on Social Media? I would love to hear from you.

Takis Metaxas

Misinformation and Propaganda in Cyberspace

Since the early days of the discipline, Computer Scientists have always been interested in developing environments that exhibit well-understood and predictable behavior. If a computer system were to behave unpredictably, then we would look into the specifications and we would, in theory, be able to detect what went wrong, fix it, and move on. To this end, the World Wide Web, created by Tim Berners-Lee, was not expected to evolve into a system with unpredictable behavior. After all, the creation of WWW was enabled by three simple ideas: the introduction of the URL, a globally addressable system of files, the HTTP, a very simple communication protocol that allowed a computer to request and receive a file from another computer, and the HTML, a document-description language to simplify the development of documents that are easily readable by non-experts. Why, then, in a few years did we start to see the development of technical papers that included terms such as “propaganda” and “trust“?

Soon after its creation the Web began to grow exponentially because anyone could add to it. Anyone could be an author, without any guarantee of quality. The exponential growth of the Web necessitated the development of search engines (SEs) that gave us the opportunity to locate information fast. They grew so successful that they became the main providers of answers to any question one may have. It does not matter that several million documents may all contain the keywords we were including in our query, a good search engine will give us all the important ones in its top-10 results. We have developed a deep trust in these search results because we have so often found them to be valuable — or, when they are not, we might not notice it.

As SEs became popular and successful, Web spammers appeared. These are entities (people, organizations, businesses) who realized that they could exploit the trust that Web users placed in search engines. They would game the search engines manipulating the quality and relevance metrics so as to force their own content in the ever-important top-10 of a relevant search. The Search Engines noticed this and a battle with the web spammers ensued: For every good idea that search engines introduced to better index and retrieve web documents, the spammers would come up with a trick to exploit the new situation. When the SEs introduced keyword frequency for ranking, the spammers came up with keyword stuffing (lots of repeating keywords to give the impression of high relevance); for web site popularity, they responded with link farms (lots of interlinked sites belonging to the same spammer); in response to the descriptive nature of anchor text they detonated Google bombs (use irrelevant keywords as anchor text to target a web site); and for the famous PageRank, they introduced mutual admiration societies (collaborating spammers exchanging links to increase everyone’s PageRank). In fact, one can describe the evolution of search results ranking technology as a response to Web spamming tricks. And since for each spamming technique there is a corresponding propagandistic one, they became the propagandists of cyberspace.

Around 2004, the first elements of misinformation around elections started to appear, and political advisers recognized that, even though the Web was not a major component of electoral campaigns at the time, it would soon become one. If they could just “persuade” search engines to rank positive articles about their candidates highly, along with negative articles about their opponents, they could convince a few casual Web users that their message was more valid and get their votes. Elections in the US, after all, often depend in a small number of closely contested races.

Search Engines have certainly tried hard to limit the success of spammers, who are seen as exploiting this technology to achieve their goals. Search results were adjusted to be less easily spammable, even if this meant that some results were hand-picked rather than algorithmically produced. In fact, during the 2008 and the 2010 elections, searching  the Web for electoral candidates would yield results that contained official entries first: The candidate’s campaign sites, the office sites, and wikipedia entries topped the results, well above even well-respected news organizations. The embarrassment of being gamed and of the infamous “miserable failure” Google bomb would not be tolerated.

Around the same time we saw the development of the Social Web, networks that allow people connect, exchange ideas, air opinions, and keep up with their friends. The Social Web created opportunities both for spreading political (and other) messages, but also misinformation through spamming. In our research we have seen several examples of propagation of politically-motivated misinformation. During the important 2010 Special Senatorial election in MA, spammers used Twitter in order to create a Google bomb that would bring their own messages to the third position of the top-10 results by frequently repeating the same tweet. They also created the first Twitter bomb targeting individuals interested in the MASEN elections with misinformation about one of the candidates, and created a pre-fab Tweet factory imitating a grass-roots campaign, attacking news organizations and reporters (a technique known as “astroturfing“).

Like propaganda in society, spam will stay with us in cyberspace. And as we increasingly look to the Web for information, it is important that we are able to detect misinformation. Arguably, now is the most important time for such detection, since we do not currently have a system of trusted editors in cyberspace like that which has served us well in the past (newspapers, publishers, institutions). What can we do?

*

Retweeting reveals communities of likely-minded people: There are 2 larger groups that naturally arise when one considers the retweeting patterns of those tweeting during the 2010 MA special election. Sampling reveals that the smaller contains liberals and the larger conservatives. The larger one appears to consist of 3 different subgroups.

Some promising research in social media has shown potential in using technology to detect astroturfing. In particular, the following rules hold true most (though not all) of the time:

1. The credibility of the information you receive is related to the trust you have towards the original sender and to those who retweeted it.
2. Not only do Twitter friends (those that you follow) reveal a similarly-minded community, their retweeting patterns make these communities stronger and more visible.
3. While both truths and lies propagate in cyberspace, lies have shorter life-spans and are questioned more often.

While we might prefer an automatic way of detecting misinformation with the use of algorithms, this will not happen. Citizens of cyberspace must become educated about how to detect misinformation, be provided with tools that will help them question and verify information, and draw on the strengths of crowd sourcing through their own groups of trusted editors. This Berkman conference will help us push in this direction.