Logic-level Evidence Retrieval and Graph-based Verification Network for Table-based Fact Verification

TL;DR

This paper introduces LERGV, a novel framework for table-based fact verification that retrieves logic-level evidence and employs graph-based reasoning to improve accuracy, addressing issues with spurious programs in prior methods.

Contribution

The paper proposes a new evidence retrieval and graph-based verification approach that enhances logical reasoning in table-based fact verification tasks.

Findings

Effective on TABFACT benchmark

Improves logical reasoning accuracy

Addresses spurious program issues

Abstract

Table-based fact verification task aims to verify whether the given statement is supported by the given semi-structured table. Symbolic reasoning with logical operations plays a crucial role in this task. Existing methods leverage programs that contain rich logical information to enhance the verification process. However, due to the lack of fully supervised signals in the program generation process, spurious programs can be derived and employed, which leads to the inability of the model to catch helpful logical operations. To address the aforementioned problems, in this work, we formulate the table-based fact verification task as an evidence retrieval and reasoning framework, proposing the Logic-level Evidence Retrieval and Graph-based Verification network (LERGV). Specifically, we first retrieve logic-level program-like evidence from the given table and statement as supplementary evidence for the table. After that, we construct a logic-level graph to capture the logical relations between entities and functions in the retrieved evidence, and design a graph-based verification network to perform logic-level graph-based reasoning based on the constructed graph to classify the final entailment relation. Experimental results on the large-scale benchmark TABFACT show the effectiveness of the proposed approach.