Skip to main content

Abstract

Misinformation has always existed in society. Nowadays, the technological development and the appearance of social networks, pseudo-newspapers, and blogs, have aggravated this problem by facilitating the rapid spread of websites with malicious intentions. This fact makes it easier to use disinformation as an attack vector for huge communities. This has led to the development of procedures that detect the appearance of these attacks and mitigate their influence. Thus, the development of software solutions able to support human experts to d etect suspicious behaviors and misleading information is currently a key proposal. These systems should gather textual content from reliable sources extracting semantic knowledge and the opinion polarity. For instance, some solutions automatize the storage, features extraction, knowledge inference, processing, and narratives comparison from multiple media sources in a single flow. The usage of machine learning models as BERT or other similar neural networks able to capture patterns from texts, and the organization of the knowledge through semantic networks or graphs are some of the most common approaches to address the issue. These tools must include ad-hoc similarity measures with the purpose of detecting the suspicious information according to specific features that univocally promote its identification.