Developed within the sphere of Horizon2020 IT research projects, Fandango brings together important international partners, including the Italian Engineering – leader – and the ANSA news agency.
Thanks to its ability to develop and manage enabling technologies, Engineering – one of the leading global companies supporting public and private companies in Digital Transformation – together with ANSA, the first news agency in Italy and among the most important news agencies in the world - presented the Fandango project on “Augmented journalism: the future of information in times of Artificial Intelligence", as part of an event organized by the news agency. Developed within the sphere of Horizon2020 IT research projects, and based on a proposal for studying Big Data and Artificial Intelligence applied to fake news presented by Engineering, Fandango has brought together important international partners (Live Tech, CERTH, Siren Solution, VRT, Civio, Polytechnic University of Madrid) meeting in a consortium, with the active participation of ANSA, which immediately collaborated as the only European agency.
The project, which started last year and is expected to be completed by the end of 2020, has the final objective of developing an IT system which, thanks to the use of the most advanced Big Data applications, Artificial Intelligence, image recognition and text analysis, can help fight fake news, promoting truthful information and the consequent benefits at economic, social and cultural level.
IT tools and services designed to support the world of media in the fast analysis of written content, audio, images and videos aim to be the end point, providing indexes of news reliability generated through the activation of Artificial Intelligence algorithms, helping information professionals in their analyses. In fact, it should be stressed that, in the light of the structured model that contributes to feeding the information system (facts, opinions, news also generated by the mixing of facts and opinions of the protagonists of the world of politics, economy, culture, etc.) and the growing difficulty in developing ever more numerous and global sources, journalistic professionalism will become even more important in its capacity for criticism and representation of reality in order to guide public opinion in the process of defining its own ideas.
In the era of social networks and high technological evolution, which can feed fake news and deep fake news, the issue is extremely relevant in Europe, as well as in the rest of the world, due to the significant distortive impact they can have on major political and social challenges as well as on business and industry.
During the presentation, Francesco Saverio Nucci, Engineering's Director of Application Research, added: “The conflict between true and false is an eternal question that no machine or algorithm can solve definitively but, if used as a contribution combined with people's professionalism, it can have a decisive impact. The Artificial Intelligence that animates Fandango aims to act as a strong ally of the press in the fight against fake news for information that is transparent, truthful and built on the basis of data”.
The Fandango project had its first verification in the European Commission on September 12 and is now being developed through the analysis aimed at three specific issues of great importance and social impact at the global level: immigration, climate change and European policies.
Different types of analysis of multiple data sources are combined on these issues:
- algorithms able to recognize and identify the relations that exist within the text and the headline of news, in the sentences and in the relations that are established between words and their frequency;
- semantic analysis for cross-checking the various information sources and authors, in order to verify the reliability of the facts contained in a news report. Furthermore, events and information (data included) cited within the text will be verified;
- multimedia analysis for identifying the connections that exist between audio content, images and videos and the topics covered in the article in question.
- analysis of images and videos, for verifying whether they have been manipulated and modified, also in relation to the so-called deep-fake phenomenon.
These analyses investigate the type of digital traces and related developments that any tampering of audio, video or graphic photo leaves; the processes of propagation and malicious dissemination; and finally, naturally the intersection of information deriving from analysis of huge open data archives. All this, together with syntactical analyses, evaluations on polarized or non-polarized news, constant comparison with similar news and much more, contributes to construction of the system and services developed by Fandango.