In recent years, technological advances and the massive use of the Internet and social networks have aggravated the problem of disinformation due to the speed at which the content is disseminated.

In this context, there are different entities specialized in the verification process (fact-checking) and they are dedicated to the monitoring and contrast of content, data and information that circulate through the network.  However, it is a slow and manual process that requires time and resources.

Therefore, it is essential to test and validate innovative technology solutions that will help optimise and automate the fact-checking process and thus help reduce the spread and impact of disinformation on society.

Polytechnic University of Madrid (in Spanish: UPM) is one of the winners on the Call for Solution: Tech against Disinformation .

The solution

The winning solution is a proposal from the AI+DA (Applied Intelligence and Data Analysis) research group of the Polytechnic University of Madrid (UPM). It is one of the winners of the Tech Against Disinformation challenge: How can technology optimise fact-checking processes and help the fight against disinformation? launched by Digital Future Society in collaboration with the verification agencies Efe Verifica, Verificat and Newtral and CCMA.
UPM’s proposal combines multilingual natural language processing (NLP) based on Artificial Intelligence, social network analysis and monitoring of information on social networks for the automated detection, monitoring and analysis of false claims.

The solution implements a multilingual architecture recognizing languages such as: Spanish, English, Catalonian, Basque or Galician, among many other international languages. This allows to compare information in different languages without translation (i.e., a tweet written in Catalan with a fact verified by an English fact-checker).​

DisTrack is a tool to help monitoring the presence of a false claim in social networks such as Twitter, and to analyse the whole spreading cascade, from its first appearance in the OSN to the last tweets published about the topic.
The proposal consists of testing the tool (DisTrack), based on AI and Natural Language Processing for the automated detection, monitoring and analysis of false statements or fake news on social networks (on Twitter).

The idea is to provide an instrument to explore the propagation cascade of a piece of misinformation circulating throughout Twitter, one of the social networks where misinformation is more present. The operation of the tool starts by inserting a new claim by the user. Then, a series of modules, including advanced AI tools, the tool will provide a graphic representation of the tweets spreading that misinformation through graphs and other instruments. Additionally, the tool will also allow to explore the interaction with those who oppose to the false claim, including fact-checkers. 

The pilot

The pilot is now in the execution phase (January 2022 – January 2023). It aims to provide a useful instrument (DisTrack tool) to analyse the spreading of a hoax that has been circulating on Twitter through a graphical interactive interface. By using DisTrack, it will be possible to analyse all interactions between users that contribute to dissemination of the false claim, together with anonymised individual information for each user, including the number of followers or following users. All this information, presented through a practical interactive interface, will instrumentalize a tool to understand the origin, the spreading process and the status of a hoax or rumour.  

The tool is being tested with the collaborating verification agencies, Newtral and Verificat, who will also evaluate the solution.

Expected outcomes

The pilot objective is to test DisTrack tool that:

– Provides a visualization of the propagation cascade of hoaxes in social media

– Detects disinformation spreaders and influencers​

– Supports tracking the origin of hoaxes​

– Helps understanding the nature of disinformation and how to tackle it