A fake news detection solution in social media via crowdsourcing, blockchain technology, and artificial intelligence

Abstract

With the exacerbation of the problem of fake news since the start of 2020's global pandemic, the authenticity of data and information now bears a greater weight with the new age of social distancing where the Internet became the standard news source to many. Data-driven and human intelligence-based approaches to tackle fake news have both seen a rise in adoption but are coupled with their respective flaws, which led to the birth of a proposed solution that combines both to identify the authenticity of news more accurately. The proposed solution utilizes Artificial Intelligence (AI) in the form of a Bidirectional Encoder Representations from Transformers (BERT) text classifier with the help of crowd intelligence to derive a novel trust index algorithm for evaluating news authenticity. The AI serves to evaluate news from a structural perspective whereas the crowd intelligence offsets the AI's potential misevaluation by performing manual fact checks based on the majority rule. Both components complement one another by toning down the other's flaws, which ultimately results in a more trustworthy outcome.