An official website of the United States government
Retrieving evidence to support or refute claims is a core part of automatic fact-checking. Prior work makes simplifying assumptions in retrieval that depart from real-world use cases: either no access to evidence, access to evidence curated by a human fact-checker, or access to evidence published after a claim was made. In this work, we present the first realistic pipeline to check real-world claims by retrieving raw evidence from the web. We restrict our retriever to only search documents available prior to the claim’s making, modeling the realistic scenario of emerging claims. Our pipeline includes five components: claim decomposition, raw document retrieval, fine-grained evidence retrieval, claim-focused summarization, and veracity judgment. We conduct experiments on complex political claims in the ClaimDecomp dataset and show that the aggregated evidence produced by our pipeline improves veracity judgments. Human evaluation finds the evidence summary produced by our system is reliable (it does not hallucinate information) and relevant to answering key questions about a claim, suggesting that it can assist fact-checkers even when it does not reflect a complete evidence set.
more »
« less
Generating Literal and Implied Subquestions to Fact-check Complex Claims
Verifying political claims is a challenging task, as politicians can use various tactics to subtly misrepresent the facts for their agenda. Existing automatic fact-checking systems fall short here, and their predictions like "half-true" are not very useful in isolation, since it is unclear which parts of a claim are true or false. In this work, we focus on decomposing a complex claim into a comprehensive set of yes-no subquestions whose answers influence the veracity of the claim. We present CLAIMDECOMP, a dataset of decompositions for over 1000 claims. Given a claim and its verification paragraph written by fact-checkers, our trained annotators write subquestions covering both explicit propositions of the original claim and its implicit facets, such as additional political context that changes our view of the claim's veracity. We study whether state-of-the-art pre-trained models can learn to generate such subquestions. Our experiments show that these models generate reasonable questions, but predicting implied subquestions based only on the claim (without consulting other evidence) remains challenging. Nevertheless, we show that predicted subquestions can help identify relevant evidence to fact-check the full claim and derive the veracity through their answers, suggesting that claim decomposition can be a useful piece of a fact-checking pipeline.
more »
« less
- PAR ID:
- 10380029
- Date Published:
- Journal Name:
- Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP)
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Textual entailment models are increasingly applied in settings like fact-checking, presupposition verification in question answering, or summary evaluation. However, these represent a significant domain shift from existing entailment datasets, and models underperform as a result. We propose WiCE, a new fine-grained textual entailment dataset built on natural claim and evidence pairs extracted from Wikipedia. In addition to standard claim-level entailment, WiCE provides entailment judgments over sub-sentence units of the claim, and a minimal subset of evidence sentences that support each subclaim. To support this, we propose an automatic claim decomposition strategy using GPT-3.5 which we show is also effective at improving entailment models’ performance on multiple datasets at test time. Finally, we show that real claims in our dataset involve challenging verification and retrieval problems that existing models fail to address.more » « less
-
Knowledge graph has been widely used in fact checking, owing to its capability to provide crucial background knowledge to help verify claims. Traditional fact checking works mainly focus on analyzing a single claim but have largely ignored analysis on the semantic consistency of pair-wise claims, despite its key importance in the real-world applications, e.g., multimodal fake news detection. This paper proposes a graph neural network based model INSPECTOR for pair-wise fact checking. Given a pair of claims, INSPECTOR aims to detect the potential semantic inconsistency of the input claims. The main idea of INSPECTOR is to use a graph attention neural network to learn a graph embedding for each claim in the pair, then use a tensor neural network to classify this pair of claims as consistent vs. inconsistent. The experiment results show that our algorithm outperforms state-of-the-art methods, with a higher accuracy and a lower variance.more » « less
-
null (Ed.)We extend evidence-aware claim verification to the context of positive-unlabeled (PU) learning. Existing works assume the truth and the falsity of the claims are known for training and form the task as a supervised learning problem. However, this assumption underestimates the difficulty of collecting false claims; we argue that claim verification is more challenging in the absence of negative labels. We consider a more practical setting, where only a comparatively small number of true claims are labeled and more claims remain unlabeled. Thus, we formulate the claim verification task as a PU learning problem. We decouple learning representation of claim-evidence pair from PU learning and adopt a pre-trained universal language model to encode claim-evidence pairs. We further propose to use the generative adversarial network (GAN) to capture the latent alignment between encoded claim-evidence pair and the truthfulness. We leverage the verification as part of the GAN by extending previous GAN based PU learning. We show that the proposed model achieves the best performance with a small amount of labeled data and is robust to the truthfulness prior estimation. We conduct a thorough analysis of the model selection. The proposed approach performs the best under two practical scenarios: (i) the unlabeled data is more than the labeled data; (ii) and the unlabeled positive data is more than the unlabeled negative data.more » « less
-
Recognizing if LLM output can be grounded in evidence is central to many tasks in NLP: retrieval-augmented generation, summarization, document-grounded dialogue, and more. Current approaches to this kind of fact-checking are based on verifying each piece of a model generation against potential evidence using an LLM. However, this process can be very computationally expensive, requiring many calls to a model to check a single response. In this work, we show how to build small fact-checking models that have GPT-4-level performance but for 400x lower cost. We do this by constructing synthetic training data with GPT-4, which involves creating realistic yet challenging instances of factual errors via a structured generation procedure. Training on this data teaches models to check each fact in the claim and recognize synthesis of information across sentences. For evaluation, we unify datasets from recent work on fact-checking and grounding LLM generations into a new benchmark, LLM-AggreFact. Our best system MiniCheck-FT5 (770M parameters) outperforms all systems of comparable size and reaches GPT-4 accuracy. We release LLM-AggreFact, code for data synthesis, and models.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
