Fake news can significantly influence this year’s elections in Brazil, according to 60% of respondents to a survey conducted by Datafolha, which involved 2,556 people over 16 years old, residing in 181 cities. Among younger people, who are more accustomed to using social networks, the negative perception about the circulation of fake news is even greater.
“Although many believe it is a difficult issue to circumvent in times of elections, fake news can indeed be reduced with the help of technology, both in terms of the barriers adopted by the platforms and the intelligent resources implemented to combat critical snippets of information,” says Marcelo Comite, VP for Latin America, of Voyager Labs, an Israeli startup specialized in investigation solutions based on Artificial Intelligence.
The company has developed an algorithm whose mission is to assist in investigations of content and data circulating on the internet and social networks, called Voyager Analytics. The platform has been enabling investigators to uncover social hangouts and hidden connections between dangerous individuals and groups by focusing on the most relevant leads and critical snippets of data, scattered in an ocean of unstructured data.
“Every day, an immense amount of data is produced on the so-called deep web – a layer of the web that is not indexed by search engines and remains hidden from the general public. Since it cannot be accessed through search engines like Google, and also because it is not accessed by typing an address in a common browser (Chrome, Firefox, Edge, etc.), criminals use the difficulty of access to share illegal content such as drug sales, pedophilia and violence. And in elections, the process of shooting fake news is no different,” says the expert.
According to him, Voyager’s solution relies on AI to analyze large amounts of open and deep data in this murky online space, as well as internal data – all with the goal of uncovering actionable insights that support investigations. “To get to the target, you have to simplify the collection of information and its analysis, which enables intelligent visualization. “This would take months to do with traditional tools,” emphasizes Comité.
In addition, the technology presents the most relevant and important information in near-real time, saving resources normally undertaken in retrieving, processing and analyzing large amounts of unstructured data. “By extracting data insights from a wide variety of sources, the tool enables its use in enriching existing leads and, consequently, in discovering new ones,” concludes Comité.