Strategies to Use Harvesters in Trustworthy Fake News Detection Systems

Today, detecting fake news is an important challenge facing societies and scientific communities using AI and IT solutions. Many previous works tackle the problem of visual content or text analysis. Still, those seldom tackle the real-life data collection and harvesting problem, mainly focusing on well-crafted and prepared old datasets. That is why, in this paper, we focus on evaluating the performance of the distributed harvester in the context of fake news detection. To this aim, we developed a test-bed that simulated a real-life scenario and conducted experiments to determine the efficacy of the harvester system. The results obtained show that the investigated problem can become a crucial issue, and thus, various aspects must be considered to address it.