Automatic Fake News Detection: Are Models Learning to Reason?
This reveals a critical flaw in current fake news detection methods, questioning their reliance on reasoning and impacting researchers and practitioners in misinformation mitigation.
The paper investigates whether fact-checking models for fake news detection actually learn to reason by analyzing the roles of claims and evidence, finding that using only evidence often yields the highest effectiveness, with claim inclusion being negligible or harmful.
Most fact checking models for automatic fake news detection are based on reasoning: given a claim with associated evidence, the models aim to estimate the claim veracity based on the supporting or refuting content within the evidence. When these models perform well, it is generally assumed to be due to the models having learned to reason over the evidence with regards to the claim. In this paper, we investigate this assumption of reasoning, by exploring the relationship and importance of both claim and evidence. Surprisingly, we find on political fact checking datasets that most often the highest effectiveness is obtained by utilizing only the evidence, as the impact of including the claim is either negligible or harmful to the effectiveness. This highlights an important problem in what constitutes evidence in existing approaches for automatic fake news detection.