HYBRIDS

César González-Pérez (Tecnoloxías Semánticas)

False rumours, fake news or hate speech against vulnerable minorities spread quickly through social media, undermining the credibility of social and political actors in both consolidated and unconsolidated democracies. The economy of countries is also harmed by the spread of false information, affecting the stock exchange shares or expectations on big investments. The dissemination of misleading and inaccurate information is facilitated due to the new social technologies. A recent study published in Science (Vosoughi 2018) showed that false rumours and misinformation diffused significantly farther, faster, deeper, and more broadly than the truth. This study also shows that false political news is the fastest and most globally disseminated news, surpassing other categories of information, such as urban legends, natural disasters, terrorism, or financial information. False rumours through social media are becoming one of the main threats against good governance and the appropriate management of public affairs by public institutions in democracies. This is especially relevant in connection to cultural heritage.

This project aims to train PhD students to address these issues through a H2020-MSCA-ITN-2020 grant.

The general objective of the project is to train a pool of specialists in innovative approaches for "hybrid intelligence". By hybrid in this context we refer to a combination of the usual natural language processing techniques with an account of social and linguistic knowledge.

Two specific objectives will be pursued:

Good Governance: The proposed ETN will develop the necessary methodologies and technologies to be able to better understand the current situation regarding threats to current governance systems (this would include the aspect of self-reflection), as well as to recommend practices and policies for action. This would include topics such as the detection and analysis of these threats, and the study of public discourse regarding namely manipulation and distortion of that discourse.

Hybrid Intelligence: The ETN will integrate the knowledge of social and human sciences specialists in ML/DL systems and NLP tools to develop new hybrid intelligence systems. This is a methodological objective. The main aim is to inject deep and structured knowledge (not just annotated text) into ML/DL and NLP systems so as to develop hybrid approaches.

Fake news. Natural language processing. Discourse analysis.

01/09/2022 - 30/08/2026

Contribución

Público competitivo