Evidence-Based Temporal Fact Verification

TL;DR

This paper tackles the challenge of verifying temporal claims by decomposing claims and supporting evidence into event-level units with explicit temporal expressions. It introduces TACV, a temporal-aware verification framework that encodes events with a sinusoidal positional temporal encoding, propagates information through a graph attention network, and leverages an LLM for per-event temporal reasoning on top-k retrieved evidence. The authors also curate two temporal datasets, T-FEVER and T-FEVEROUS, and demonstrate that TACV substantially outperforms state-of-the-art baselines on temporal and real-world datasets, while providing interpretable event-level reasoning. The work advances automated temporal fact verification and offers resources and insights for robust time-sensitive information verification in real-world applications.

Abstract

Automated fact verification plays an essential role in fostering trust in the digital space. Despite the growing interest, the verification of temporal facts has not received much attention in the community. Temporal fact verification brings new challenges where cues of the temporal information need to be extracted and temporal reasoning involving various temporal aspects of the text must be applied. In this work, we propose an end-to-end solution for temporal fact verification that considers the temporal information in claims to obtain relevant evidence sentences and harness the power of large language model for temporal reasoning. Recognizing that temporal facts often involve events, we model these events in the claim and evidence sentences. We curate two temporal fact datasets to learn time-sensitive representations that encapsulate not only the semantic relationships among the events, but also their chronological proximity. This allows us to retrieve the top-k relevant evidence sentences and provide the context for a large language model to perform temporal reasoning and outputs whether a claim is supported or refuted by the retrieved evidence sentences. Experiment results demonstrate that the proposed approach significantly enhances the accuracy of temporal claim verification, thereby advancing current state-of-the-art in automated fact verification.

Figures (3)

• Figure 1: Overview of TACV framework.
• Figure 2: Temporal-aware representation encoding.
• Figure 3: Effect on $k$ on TACV. Solid line depict label accuracy, dotted line depict FEVER score.