Give Me More Details: Improving Fact-Checking with Latent Retrieval

TL;DR

This paper introduces a latent variable model that leverages full source documents for fact-checking, improving evidence extraction and claim verification beyond traditional snippet-based methods.

Contribution

It proposes a novel latent retrieval approach that jointly extracts evidence and verifies claims using full text documents, along with new multilingual and monolingual datasets.

Findings

Including source documents enhances evidence retrieval.

The model outperforms existing methods in claim verification.

Significant improvements are achieved across various settings.

Abstract

Evidence plays a crucial role in automated fact-checking. When verifying real-world claims, existing fact-checking systems either assume the evidence sentences are given or use the search snippets returned by the search engine. Such methods ignore the challenges of collecting evidence and may not provide sufficient information to verify real-world claims. Aiming at building a better fact-checking system, we propose to incorporate full text from source documents as evidence and introduce two enriched datasets. The first one is a multilingual dataset, while the second one is monolingual (English). We further develop a latent variable model to jointly extract evidence sentences from documents and perform claim verification. Experiments indicate that including source documents can provide sufficient contextual clues even when gold evidence sentences are not annotated. The proposed system is able to achieve significant improvements upon best-reported models under different settings.