An Aid to Mitigate Online Misinformation with Self-Sovereign Identity and Artificial Intelligence

The issue of misinformation in Online Social Networks (OSNs) has been a persistent challenge traditionally addressed through automated artificial intelligence tools and human fact-checkers. However, existing solutions do not assess the credibility of a post based on the author’s competence in the subject matter. This work presents an approach to fill this gap by leveraging emerging Self-Sovereign Identity (SSI) technology to generate verifiable competence certificates for content authors. Our solution combats misinformation in OSNs by offering a new perspective on content credibility. SSI is a paradigm of digital identity that enables individuals to present cryptographically provable certificates, known as Verifiable Credentials (VCs), which attest to their attributes, such as skills and qualifications. We propose a system in which authors attach VCs to their posts, providing readers with verifiable proof of the authors’ competence regarding their posts’ topics and thus enhancing trust in the authors. To match skills attested in the VCs with post content, we employed a system equipped with a Named Entity Recognition (NER) model to extract keywords from posts and map them to an ontology of skills and competencies. Experimental results on a synthetic dataset generated using GPT-40 demonstrate that our approach effectively reduces the spread of misinformation in a simulated OSN environment. We strongly believe that integrating verifiable author expertise attestations enhances the credibility of posts.