This early-stage PhD paper introduces the intended research of the author’s PhD study. The research is focused on explaining misleading claims, which have been fact-checked, using graphs of entities. This approach combines argumentation modeling with the fact-checking domain. Existing work seems to focus more on whole sentences concerning argumentation, while this research aims to go deeper at the level of entities and relationships between them. Preliminary studies have already revealed some repetitive argumentation elements that are based on an entity graph. The results from this study should help us understand the misleading claims better and suggest how knowledge behind unstructured data (meaning texts) can be expressed in formal representation.