GEAR: Graph-based Evidence Aggregating and Reasoning for Fact Verification

By Jie Zhou, Xu Han, Cheng Yang, Zhiyuan Liu, Lifeng Wang, Changcheng Li, Maosong Sun. 2019

241 citations recorded by the source corpus.

Fact verification (FV) is a challenging task which requires to retrieve relevant evidence from plain text and use the evidence to verify given claims. Many claims require to simultaneously integrate and reason over several pieces of evidence for verification. However, previous work employs simple models to extract information from evidence without letting evidence communicate with each other, e.g., merely concatenate the evidence for processing. Therefore, these methods are unable to grasp sufficient relational and logical information among the evidence. To alleviate this issue, we propose a graph-based evidence aggregating and reasoning (GEAR) framework which enables information to transfer on a fully-connected evidence graph and then utilizes different aggregators to collect multi-evidence information. We further employ BERT, an effective pre-trained language representation model, to improve the performance. Experimental results on a large-scale benchmark dataset FEVER have demonstrated that GEAR could leverage multi-evidence information for FV and thus achieves the promising result with a test FEVER score of 67.10%. Our code is available at https://github.com/thunlp/GEAR.

Subfields: Artificial Intelligence, Management Science and Operations Research, Natural Language Processing, Graph Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Management Science and Operations Research

In Natural Language Processing

In Graph Learning

Citations in the corpus

Answering Complex Open-Domain Questions with Multi-Hop Dense Retrieval (2020, background): “…C for the exact numbers. the results in veriﬁcation label accuracy (LA) and the FEVER score 9 in Table 7, where the numbers of competitive baselines, GEAR (Zhou et al., 2019), graph attention network (GAT) (Veliˇckovi´c et al., 2017) and variants of KGAT are from the KGAT (Liu et al., 2020) paper.”
Fine-grained Fact Verification with Kernel Graph Attention Network (2019, methodology): “The document retrieval step retrieves related Wikipedia pages and is kept the same with previous work (Hanselowski et al., 2018; Zhou et al., 2019; Soleimani et al., 2019).”
A Multi-Level Attention Model for Evidence-Based Fact Checking (2021, methodology): “GEAR (Zhou et al., 2019), KGAT (Liu et al., 2020), and DREAM (Zhong et al., 2020) are graph-based models.”
Transformer-XH: Multi-Evidence Reasoning with eXtra Hop Attention (2020, methodology): “Many previous FEVER systems first retrieve the evidence sentences for the claim and then reason verify it (Nie et al., 2019; Zhou et al., 2019).”
Graph Sequential Network for Reasoning over Sequences (2020, methodology): “Both tasks require the model to have reasoning ability, and top performance has been achieved with early summarization followed by GNN (Zhou et al., 2019; Xiao et al., 2019; Fang et al., 2019; Tu et al., 2019a).”
Kernel Graph Attention Network for Fact Verification (2019, methodology): “GEAR (Zhou et al., 2019) reasons and aggregates over claim evidence pairs with Graph Attention Network (GAT) (Veliˇckovi´c et al., 2017).”
Fool Me Twice: Entailment from Wikipedia Gamification (2021, background): “For the second component of the task, we follow state-of-the-art entailment models (Zhou et al., 2019; Liu et al., 2020; Eisenschlos et al., 2020):”
Are Factuality Checkers Reliable? Adversarial Meta-evaluation of Factuality in Summarization (2021, methodology): “, first retrieval then verification (Zhou et al., 2019)) should be put forwards to deal with the problem.”
DialFact: A Benchmark for Fact-Checking in Dialogue (2021, methodology): “methods for document retrieval: The first one is WikiAPI5, which retrieves Wikipedia pages and is used in past fact-checking work (Hanselowski et al., 2018; Stammbach and Neumann, 2019; Zhou et al., 2019; Liu et al., 2020).”
Program Enhanced Fact Verification with Verbalization and Graph Attention Network (2020, background): “Others formulates fact veriﬁcation as graph reasoning (Zhou et al., 2019; Liu et al., 2020).”
Coreferential Reasoning Learning for Language Representation (2020, methodology): “Baselines We compare our model with four BERT-based fact veriﬁcation models: (1) BERT Concat (Zhou et al., 2019) concatenates all evidence pieces and the claim to predict the claim label; (2) SR-MRS (Nie et al., 2019) employs hierarchical BERT retrieval to improve model performance; (3) GEAR (Zhou e”
DTCA: Decision Tree-based Co-Attention Networks for Explainable Claim Verification (2020, background): “…discovery of evidence for explainable claim veriﬁcation, which mainly designs different deep models to exploit semantic matching (Nie et al., 2019; Zhou et al., 2019), semantic conﬂicts (Baly et al., 2018b; Dvo ˇ r ´ ak and Woltran, 2019; Wu and Rao, 2020), and semantic entailments (Hanselowski…”
SimGRAG: Leveraging Similar Subgraphs for Knowledge Graphs Driven Retrieval-Augmented Generation (2024, unknown): “For fact verification, the KG version of GEAR (Zhou et al., 2019) is trained on the FactKG training set.”