Fake News — An Overview

Fake News – An Overview

“Fake news” has become a frequently used term in the past few years to describe a phenomenon specific to online social media platforms like Facebook, or Twitter. Through functions such as Posts, Likes or Shares, entities and persons can distribute content which accords with more or less severe characteristics of the term “fake news”. The Royal Statistical Society in Great Britain provides the following definitions:

  • Entirely made up news resulting from the desire to make money, or from disinformation campaigns.
  • Hyper-partisan news which may have some basis in fact but paints a distortive picture
  • News which is so poor in its use of statistics and information that it is factually incorrect[1]

These definitions emphasise the different forms of impact fake news have, an impact which is also traceable in numbers: A Buzzfeed News analysis showed that during the last three months before the 2016 US election, the 20 best performing false stories about the election generated 8,711,000 shares, reactions and comments on Facebook. In contrast, the 20 best performing election from major news websites fell short of this with 7,367,000 shares, likes or comments[2].

This poses the question: Why are fake news prospering while traditional news media struggle for relevance?

A key to the answer lies in the financial functioning of social media and digital environments in general. For every online media outlet, revenue is generated through the combination of clicks (e.g. on an article) and the ads accessed along with the article. This scheme favours high click numbers independent of the accessed content[3]. In turn, content is not only produced with regard to a high user appeal, but content is also produced in high numbers to engage the audience repeatedly.

While traditional news outlets invest time and money into research and verification of facts before publishing their articles to generate revenue, fake news articles seek to attract as many people and clicks as possible for the highest possible revenue. The quality and reliability of information is abandoned in favour of appealing language, e.g. exaggerated and thus catchy headlines. The low production costs of fake news are thus rewarded with high outcome, while decreasing revenue for traditional news outlets create difficulties to maintain staff and resources for proper research and coverage. This is relevant in a surprisingly local context: newspapers and smaller publishers cut their offices in smaller, rural towns to centralize their resources in bigger cities. This creates not only a gap in their news coverage, but also to the readership left in those regions, which then have to search for new sources of information on their immediate environment[4].

Another key element is the user’s position: Facebook serves increasingly as main source for news instead of traditional news websites[5]. This is a crucial factor with regard to Facebook’s filter algorithm: other than on a website, users are no longer exposed to random content independent of their habits or political stances. Instead, on Facebook they are confronted with likeminded content and users, creating the amplifying effect of the echo-chamber[6]. This also plays into the human tendency to believe information and reports which confirm already existing beliefs[7].

The mass of content produced and increasing broadness of media outlets with a platform on Facebook blurs lines between established traditional and newcomer news outlets, as well as between reliable and unverified news. A study by British Channel4 showed that 71% of internet users with Facebook as primary source for news believe that one out of three fake news stories is true, while merely 47% of those who rely on broadcasters shared this notion[8].

Considering that every user on Facebook is thus vulnerable to the blurred lines between verified and fake news, the question comes up: How is it possible to distinguish fake news from verified articles? Here are some key markers.

The Post

The appeal of the post on Facebook is often founded on aspects of the so-called “clickbait” headline[9]:

  • Excessive length of ten or more words
  • Contractions usually restricted to spoken language:
    “you’ll”, “you’re”, “they’re”
  • Emotive, overly positive or negative adjectives:
    “breath-taking”, “awe-inspiring”, “horrible”
  • Curiosity-raising phrases:
    “will blow your mind”, “you won’t believe”
  • Forwarding language referring to further content only to be found in the article:
    “this”, “my”, “which”, “here”
  • Numbers demonstrating the dimension and certainty of the content:
    “WHOA: 4 Questions That Got 120 Rapists To Admit They Were Rapists”
  • Vague language, also called “engines” for their curiosity-raising appeal at the centre of the headline: “thing” in “Most of These People Do the Right Thing, But the Guys at the End? I Wish I Could Yell At Them.”

The Article

Clickbait headlines alone are not characteristics unique to fake news, since Rubin points out that its sensationalist and eye-catching appeal are also common with tabloid and yellow press[10]. Following indicators can help to determine whether an article or website is reliable:

Authorship & Authority

  • Author’s name or contact, alternatively an “About”-page
  • Information on their qualifications and credentials

Content – Verification

  • Language:
    • Correct grammar, spelling and typography
    • Avoidance of spoken, colloquial language
  • Date of first publication and revision
  • Links referring to article’s sources
    • Availability
    • Date of source’s publication
  • Entities referring to the article fulfil criteria of reliability
  • Index and site maps referring to previous coverage of topic

Content – Objectivity

  • Comparison
    • other reliable source’s approach to and information on the topic (institutional websites, encyclopedia, further news outlets)
  • Opinion
    • Author’s language and formulation sticks to neutral explanation
    • Lack of additional judgemental adjectives
    • Opposing positions are represented in the article, e.g. through official statements or requests for them

Technical Indicators

  • URL structure[11]:
  • Administrative type of source
    • .org = non-governmental, professional organisation
    • .gov = government website
    • .com = corporate, business, profit-based website
    • .edu = educational institution, academic
    • Indication of geographic host location: .hu, .de, .it, .ru, .uk

This list is being continuously updated.

[1] http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/culture-media-and-sport-committee/fake-news/written/48238.html

[2] https://www.buzzfeed.com/craigsilverman/viral-fake-election-news-outperformed-real-news-on-facebook?utm_term=.ojwng9ONBY#.luzWXRjpNv

[3] http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/culture-media-and-sport-committee/fake-news/written/48244.html

[4] http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/culture-media-and-sport-committee/fake-news/written/48244.html#_ftn49

[5] http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/culture-media-and-sport-committee/fake-news/written/48244.html

[6] http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/culture-media-and-sport-committee/fake-news/written/48238.html

[7] http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/culture-media-and-sport-committee/fake-news/written/48231.html#_ftn11

[8] http://www.channel4.com/info/press/news/c4-study-reveals-only-4-surveyed-can-identify-true-or-fake-news

[9] Chakraborty, Abhijnan, et al. “Stop Clickbait: Detecting and preventing clickbaits in online news media.” Advances in Social Networks Analysis and Mining (ASONAM), 2016 IEEE/ACM International Conference on. IEEE, 2016.

[10] Rubin, Victoria L., Yimin Chen, and Niall J. Conroy. “Deception detection for news: three types of fakes.” Proceedings of the Association for Information Science and Technology 52.1 (2015): 1-4.

[11] https://www.projectlooksharp.org/?action=webresources_websites