Aletheia, a Greek word for the unveiling of truth, has had alternate fortunes in the history of Western thought, becoming appreciated in the philosophical debate of the twentieth century. Today it has again become central to social and political discussion, after the Oxford English Dictionary defined and recognized the term Post-truth as “word of the year” in 2016. Since then, this theme has seen key roles being taken on by fake news, the political debate around the theme of misinformation (which we must not translate with the simple concept of disinformation, whose connotation is much more intelligence-oriented) and finally deep fakes, that is, videos similar to real ones, but elaborately fabricated, thanks to Artificial Intelligence techniques. How greatly these issues have an impact on our daily lives today is demonstrated by the fact that many operators believe they are capable of putting at risk the proper performance of democratic processes.
Climate, immigration and the European context are typical scenarios in which false news can influence the perception of social and commercial actions. These examples are related to false climate data reports, which generate confusion concerning potential climate change threats and which can influence corporate and business choices, as well as fake news about immigrants or the European context, where there is a worrying disinformation.
How can technology counteract fake news and disinformation?
We live in a world which is constantly connected and for this reason the processing time in the fight against fake news is fundamental: labeling should take place while the fake news is still being propagated “live” so as not to explode on the Web and then reach traditional media. This is why analysis, correlation and evaluation techniques of various data sources in an investigative manner are necessary, as well as modern applications of Artificial Intelligence to be used together with human intervention, which still has the last word.
In recent years, Engineering’s research and development laboratories have created Fandango, a project co-funded by the European Commission which uses Artificial Intelligence techniques not to create deep fakes or disinformation, but rather to help fight these.
One of the side effects of the explosion of digital media is the phenomenon of the abundance of data from different sources (press companies, journalists, prosumers, commentators, etc.), which makes it extremely difficult to filter and discriminate valid information from that which is intentionally incorrect and often deliberately maliciously generated. Thanks to advanced Artificial Intelligence algorithms, Fandango assists journalists, fact checkers and media professionals to identify a piece of news as correct or false through the retrieval, correlation and evaluation of large collections of different and heterogeneous data and news sources.
Is there a mechanism to tell whether an article is true or false?
Today, several media and technology companies would like to have fully automated solutions that simply mark an article as “true” or “false”. But reality is much more difficult than it may appear. It is clear that Big Data tools can help to understand the information context concerning different topics, analyzing these from different points of view and evaluating the statements attributed to the event, thus helping to raise awareness about the news item.
The automatic learning algorithms used by Fandango are able to exploit the Big Data context and Deep Learning to search for correlations within the data and to decide whether a specific piece of information is false or not, or to evaluate this information with a falsehood indicator, thus helping to assess the reliability of a particular news item.
Fandango shows how, while on the one hand digital technology has caused misinformation and fake news, on the other it is still this same digital technology which can provide useful solutions to fight these.
Francesco Saverio Nucci