Saved in:
Bibliographic Details
Main Authors: Liu, Xiangyu, Liu, Yang, Hu, Wei
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.12108
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909074734448640
author Liu, Xiangyu
Liu, Yang
Hu, Wei
author_facet Liu, Xiangyu
Liu, Yang
Hu, Wei
contents Knowledge graphs (KGs) often contain various errors. Previous works on detecting errors in KGs mainly rely on triplet embedding from graph structure. We conduct an empirical study and find that these works struggle to discriminate noise from semantically-similar correct triplets. In this paper, we propose a KG error detection model CCA to integrate both textual and graph structural information from triplet reconstruction for better distinguishing semantics. We design interactive contrastive learning to capture the differences between textual and structural patterns. Furthermore, we construct realistic datasets with semantically-similar noise and adversarial noise. Experimental results demonstrate that CCA outperforms state-of-the-art baselines, especially in detecting semantically-similar noise and adversarial noise.
format Preprint
id arxiv_https___arxiv_org_abs_2312_12108
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Knowledge Graph Error Detection with Contrastive Confidence Adaption
Liu, Xiangyu
Liu, Yang
Hu, Wei
Computation and Language
Artificial Intelligence
Knowledge graphs (KGs) often contain various errors. Previous works on detecting errors in KGs mainly rely on triplet embedding from graph structure. We conduct an empirical study and find that these works struggle to discriminate noise from semantically-similar correct triplets. In this paper, we propose a KG error detection model CCA to integrate both textual and graph structural information from triplet reconstruction for better distinguishing semantics. We design interactive contrastive learning to capture the differences between textual and structural patterns. Furthermore, we construct realistic datasets with semantically-similar noise and adversarial noise. Experimental results demonstrate that CCA outperforms state-of-the-art baselines, especially in detecting semantically-similar noise and adversarial noise.
title Knowledge Graph Error Detection with Contrastive Confidence Adaption
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2312.12108