Political turmoil in Brazil: Developing tools to combat dis/misinformation
Dis/Misinformation is harming people and institutions, representing a huge challenge for Brazil. This is especially true this year as the country will face a presidential election. In order to find solutions to this problem, prototypes were built during a hackathon in São Paulo.
Dis/misinformation are haunting social networks in Brazil. According to recent research from USP, 12 million profiles in Brazil regularly share dis/misinformation. People create rumours for political gain, profit, or other kinds of benefits. Their lies are then propagated by robots or by people who don’t understand the damage which they are causing.
More than 130 million Brazilians are active on social media and are exposed to dis/misinformation. Rumours are spreading on the web constantly. The murder of Rio de Janeiro’s city councilwoman, Marielle Franco, is an example of a case that required an enormous amount of fact-checking work: they had to prove that no, she wasn’t the spouse of a drug dealer; no, she was not supported by a faction; and no, she did not get pregnant when was teenager. Pure cases of disinformation!
Combating Dis/Misinformation is a key to press coverage in Brazil, a country which is experiencing intense political turmoil: since 2014, more than 200 politicians and businessmen went to jail as a result of the Lava Jato operation. In addition to this, a President was recently impeached and a former President, who is one of the most popular Brazilian leaders, has just been sent to prison. In this context, Brazil is heading towards a presidential election process in October.
Developing solutions at the Globo Editors Lab
With the importance of fighting against dis/information in mind, the Global Editors Network (GEN) and Globo Lab, with the support of Google News Initiative gathered some of the best Brazilian media innovators in Sao Paulo for a two-day hackathon (7–8 April) to devise new ways to fight against misinformation, disinformation, and the bots that spread them.
Ten teams participated, each of them including a journalist, a developer, and a designer, representing different media outlets (G1, Infoglobo, Zero Hora), media startups/newsrooms (Nexo, Aos Fatos, Agência Mural, Voz das Comunidades), or universities (PUC Rio, USP, Cásper Líbero).
Before working on the prototypes, the hackathon teams participated in a workshop conducted by Google News Lab Lead in Brazil Marco Túlio Pires. He gave information about the tools provided by Google to check the veracity of the content that is found on the internet.
Professor Eugenio Bucci, from the Institute of Advanced Studies of the University of São Paulo, discusses information disorders in Brazil ahead of the election period. He gave the participants historical context of the press and dis/misinformation in Brazil. He said:
‘Fake News falsifies the press’ authority. They use the natural credibility of a regular article to spread lies. They’re doing so because it’s profitable, they can gain money out of so many lies being spread.’
Marcia Menezes, from G1, talked to the hackathon teams about what newsrooms need to better fight mis- and disinformation. She reinforced the journalist’s role in the battle:
‘Journalists can’t rely on a single source or tool, the reporter has to be everywhere, has to search everywhere, and try to always use every tool available to get hold of the information.’
The prototypes to combat misinformation, disinformation, and the bots that spread them
The teams developed 10 prototypes:
Tools for journalists and researchers
BotStalker: offers tools for journalists and researchers to understand how bots will perform on Twitter during the 2018 election campaign in Brazil. The dashboard will show in real time the number of posts, the most discussed topics, the most shared links, and the most retweeted posts by the bot armies of each political group. This project will enable newsrooms to make fact-checking quicker, minimising dis/misinformation’s impact. The prototype was developed by Gaucha ZH team.
Apura Domínio: (‘Investigate Domain’) created by Nexo is a site that gathers useful public data to investigate fake news outlets. With one click, users can find the domain’s ownership, registration date, whether it has SSL certificate, as well as the Facebook pages and Twitter profiles responsible for spreading the site’s content. The project is aimed at journalists and researchers interested in investigating the production of disinformation on the web.
NewsAnalytics: the team from University of Sao Paulo created a prototype to automate the different steps of the fact-checking process to make it easier and quicker for journalists.
Ways to alert people about dis/misinformation
Trueet: (true + tweet) developed by the team from Infoglobo is an app that makes it easier for users to inform their friends about the dis/misinformation they are sharing.
MCI: the project developed by the Voz das Comunidades group wants to help people from the favela Complexo do Alemão learn how to protect themselves from dis/misinformation. They created a method to better dispel rumours among the community via WhatsApp groups created and managed by a fact-checker group.
Fact-checking in partnership with the audience
Se Liga News: created by Agencia Mural, an app designed to help users, especially less educated people, identify possible dis/misinformation on social networks. After having installed the app, users just have to copy the URL of the article and the app will tellm them whether the website can be trusted. The app also gives the users some tips to identify dis/misinformation using simple language.
G1 — é ou não é: developed by the G1 team is a bot for Facebook Messenger. Users can send the chatbot article links to find out whether they can be trusted.
No Money For A Fake is a browser extension that allow users to flag dubious URLs. This will help the team from Aos Fatos to create a blacklist of disinformation websites that they can provide to the advertising platforms that want to avoid giving ad money to disinformation websites.
Playful forms to combat dis/misinformation
Leu e duvidou?: is a media literacy website that guides the users through the fact-checking process. Developed by the team from Cásper Líbero, it provides questions to be answered by the reader, such as: ‘Did you check the domain?’, ‘Did you check the identification of the journalist who wrote the content?’ By answering the questions, the user will be able to know if the content is true or false.
PYMI: this prototype, developed by the team from Puc Rio, proposes that the users participate in bets to check if a piece of news is true or false. The user can either be rewarded or penalised for ‘betting’ if the information at hand is true or not. The idea is to make users, through the fear of losing money or the greed of making it, pay more attention to the information they are sharing.
The winning team developed a prototype which alerts users on social media about dis/misinformation to stop it from being shared
The tool developed by the Infoglobo team won the competition. The Trueet (true + tweet) tool is a Progressive Web App (PWA) that helps Twitter users debunk the false information that are spread by their network.
How does it work?
- When fact checkers identify false information on the Internet and write about them, the Trueet sends a notification to PWA users with a brief message about the false information and the link to the debunking.
- Then it scans their Twitter’s timeline and identifies, publications and retweets that are disseminating this specific false information on user’s timeline.
- Finally it allows the user to send automatically generated posts their friends, warning them in a cool way, about the false information.
Jury member Marco Tulio Pires said of the winning prototype:
‘I think they checked all the boxes we needed to check. They had a prototype that worked, they used technology in an innovative way, they took care of like, even thinking about the space in the user’s phone, which is real estate nowadays. So I really think they have all it takes to be a strong contender during the Editors Lab Final at the GEN Summit.’Jury member Evangeline de Bourgoing added:
‘Popular social networks make it difficult for people to judge the credibility of any message, because posts from publications as unlike as the New York Times and a conspiracy site look nearly identical. This means that people are increasingly reliant on friends and family members to guide them through the information ecosystem.” wrote Claire Wardle and Hossein Derakhshan . — The team from Infoglobo understood this phenomenon well and, by providing a user-friendly tool to those who want to debunk dis/misinformation among their friends, are making a promising contribution to build a healthier news ecosystem ahead of the Brazilian elections.’
The teams from Voz das Comunidades and Nexo won special mentions. The Infoglobo team will participate in the Editors Lab Final, which will take place at the GEN Summit 2018 in Lisbon, Portugal, on 30 May to 1 June 2018.
