minutes of reading

How to fact-check information?

Confronted daily with hoaxes, which are jokes or fake-news, knowing how to detect them has become indispensable. If fact-checking remains a journalistic discipline, using its methods is essential in everyday life, both in a personal and professional context. How to proceed? Buster.Ai offers you the keys.

Adopt the right mindset to detect false information

Develop your critical mind to identify false information

First of all, the discipline of fact-checking relies on our ability to develop a critical mind. We can define it as the combination of a state of mind and a set of practices that we will study together.

We need to awaken our curiosity and dedicate time to reading and analyzing information. Don’t worry, time is a precious resource, but the more we practice fact-checking, the faster we get.

Secondly, adopting a critical mind implies being aware of our biases that can alter our thinking. Yes, we have all developed biases, which, in short, are simply prejudices constructed by our minds to speed up basic cognitive operations. What we are trying to do here is to be as lucid as possible.

Asking the right questions to verify information

To help us do this, a few questions are essential when approaching a subject to be verified. What do I know? What are my assumptions? What do I not know? In short, how can we distinguish facts from interpretations?

Secondly, we must always remain modest. When faced with information, whatever its field, no one has exhaustive knowledge. Beyond this knowledge, let's accept that reality is complex, that we do not master all its mechanisms and that we can make mistakes.

Adapting to our environment

As a citizen or as a professional working for an organization, we must deal with a large flow of information on a daily basis, from a wide variety of sources: social media, newspapers or press agencies, radio or television. And let's not stop there, since a significant part of it comes from our social circles, friends, colleagues, and professional partners. This information can be of different kinds and can be of a scientific, economic, political, or social nature. Each piece of information that reaches us is likely to be totally or partially false, partly faithful, or totally true.

What is the fact-checker’s methodology?

When a piece of information, whether it is a statement, data, or a multimedia file, catches your attention because it is shocking or surprising, or because it appeals to your emotions, opinions, or convictions, take the time to check it. Note that it is only possible to inspect facts, not opinions.

As Daniel Patrick Moynihan, New York State Senator and Ambassador to India and then to the United Nations, pointed out, "Everyone is entitled to an opinion, but not entitled to their own facts."

1) Start by analyzing the source of the information

If you're on a social network, take the time to look at the form of the message:

  • When was it posted?
  • By whom?
  • In what context?
  • If audiovisual media is present, is it correlated to the content of the message?

Authentic files are regularly diverted from their initial contexts to serve unrelated causes.

Also, are there elements that allow you to affirm 100% that the image, video, or soundtrack is authentic? More importantly, does it correspond to what the author claims? To answer these questions, let's always leave room for doubt. As with a text, the more a media file plays on emotion, the more this disbelief should be pronounced. Thus, a reverse image search can be decisive. It allows to access the initial publication context and to obtain the metadata, i.e. the date of realization, the place, or the description. If fact-checking agencies have been used, they can complement your research work.

2) Find out about the author of the statement

If it is a person from your environment

Keep in mind that we tend to give more credit to people around us.

If the information is posted online 

Analyze the author's profile:

  • Is it an authentic account, a bot, a troll, or a parody account?
  • How long has it existed and what is the nature of its publications?
  • Do they follow an allegiance? Do his posts contain links?
  • What types of sites do they link to?

We should note that on Twitter, 25% of the links to news sites point to media that spread false information. 

To protect yourself from sites that impersonate media outlets, check to see if the URLs match the information being claimed. Then, let's try to determine the editorial line and the credibility of the site. Keep in mind that even an academic or institutional site can contain misinformation or disinformation.

3) Go back to the primary source to compare data

It is important to do both of these things. First, go back to the source of the claim, that is, the primary information. Is it communicated in the publication by the author? If not, be extra vigilant. 

Next, your goal should be to cross-check the claims with other sources. The goal is to test them with contradictory reasoning. To do this, you can rely on independent fact-checking agencies that have already dealt with the subject you are concerned with. If this is not the case, continue to look for primary sources: academic or scientific publications, institutional reports, etc.

Depending on the degree of difficulty, secondary sources such as the news media may be sufficient.

Are the facts presented exhaustive? 

Be sure to keep the numbers in proportion and ask yourself what the claim leaves out. We can be misled by scale effects and a partial presentation can mislead.

You now have sufficient depth of analysis based on a sufficiently broad spectrum of evidence. You can reach a verdict. To put it simply, it is defined according to the degree of veracity of the information: totally accurate, partially accurate or incomplete, partially wrong or totally wrong.

Buster.Ai, a solution for veracity. 

Fact-checker is tedious

Carrying out all the steps we have just described can be tedious, especially when it comes to making strategic and critical business or financial decisions.

It is not objectively possible for an individual to verify hundreds of statements every day using a satisfactory number of sources. Moreover, verifying information implies directing one's research toward the most relevant sources, which requires prior identification work.

The risk of error in fact-checking

Then, the risks of errors or inaccuracies exist and are not negligible. For example, how can we be sure that a report that affirms or refutes a piece of information has not itself been challenged by a more recent, methodical, and relevant study? On an individual level, the analysis can be biased by the prejudices that are unconsciously deployed by an individual in the face of cognitive over-solicitation.

Based on this observation, the Buster.Ai teams are working to develop a solution that offers both a speed of execution adapted to the mass of information to be processed and an accuracy that allows for a reliable verdict.

Buster.Ai's algorithms are true fact-checkers

To achieve this, the algorithms developed by Buster.Ai are based on deep learning artificial intelligence. The machine is able to read documents, then perform semantic analysis and text comparison. These analyses are performed with millions of sources from news agencies and titles, universities, institutions, and governments.

The verdicts are always accompanied by the most relevant passages from automatically selected sources. Thus, the verdict will validate or refute, totally or partially, the query. In detail, it can be:: 

True: the query is accurate without any important element being omitted.

Partially true: the query is partly true, but it omits important details or quotes elements out of context.

Partially false: the request contains some true elements, but ignores essential facts that may be in error.

False: the statement is inaccurate.

With Buster.Ai's artificial intelligence, the individual's ability to make informed decisions is greatly enhanced, both in the accuracy of the facts and in the depth of their understanding.

Contact our teams to discover how Buster.Ai can protect you from misinformation through practical cases.