logo
科言猫学术研究的AI总结
language
手机端手机端
  • Phone Feature微信小程序
简体中文简体中文
notice消息通知
博客博客
delete
arrow
返回

MMD-TKGR: Multi-agent Multi-round Debate for Temporal Knowledge Graph Reasoning

delete2026-01-01
delete0
PRE
AI
Y
Yadong Wu
L
Li, Xuanhong
胡泊 封面图
胡泊 (Po Hu) *
DOI：10.1007/978-981-95-3343-5_15delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
时序知识图谱推理（TKGR）旨在通过利用时序知识图谱中的历史事实来预测未来事件或关系。通过整合知识图谱的结构化语义与时间动态，TKGR非常适合涉及时间依赖推理的任务，如时序关系抽取和事件预测。近年来，大型语言模型（LLMs）在TKGR中的应用日益增多。然而，单个LLM代理通常难以有效处理复杂推理任务，这主要归因于其相对狭窄的推理视角和浅层推理深度。为应对这些挑战，本文提出了一种用于时序知识图谱推理的多代理多轮辩论框架（MMD-TKGR）。该框架通过引入关系-时间协同检索机制，优化了特定查询的历史上下文构建；同时，基于多代理协作的辩论结构动态校准预测结果并增强其可解释性。实验结果表明，所提出的方法在多个基准数据集上显著优于现有方法。
Keyword:
Temporal Knowledge Graph Reasoning
Large Language Models
Multi-Agent Debate

期刊

N
NATURAL LANGUAGE PROCESSING AND CHINESE COMPUTING, NLPCC 2025, PT I
IF:
0
论文数:
37
被引数:
0

机构

C
central china normal university
学者数:
2.8K
论文数: 1.0K
被引数: 0
W
wuhan university
学者数:
8.1W
论文数: 5.8W
被引数: 70
引用论文

引用论文

Learning Neural Ordinary Equations for Forecasting Future Links on Temporal Knowledge Graphs
errProceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
IF0
err2021-01-01
err0
errOAAI
errZhen Han; Zifeng Ding; Yunpu Ma; Yujia Gu; Volker Tresp
err分享
err收藏
ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMsReConcile：通过多样化大语言模型之间的共识，圆桌会议改进推理
errProceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
IF0
err2024-01-01
err0
PREAI
errJustin Chen; Swarnadeep Saha; Mohit Bansal
err分享
err收藏
err
IF0
err
err0
PREAI
err
err分享
err收藏
TimeTraveler: Reinforcement Learning for Temporal Knowledge Graph Forecasting
errProceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
IF0
err2021-01-01
err0
errOAAI
errHaohai Sun; Jialun Zhong; Yunpu Ma; Zhen Han; Kun He
err分享
err收藏
GenTKG: Generative Forecasting on Temporal Knowledge Graph with Large Language Models
errFindings of the Association for Computational Linguistics: NAACL 2024
IF0
err2024-01-01
err0
errOAAI
errRuotong Liao; Xu Jia; Yangzhe Li; Yunpu Ma; Volker Tresp
err分享
err收藏
HyTE: Hyperplane-based Temporally aware Knowledge Graph Embedding
errProceedings of the 2018 Conference on Empirical Methods in Natural Language Processing
IF0
err2018-01-01
err0
errOAAI
errShib Sankar Dasgupta; Swayambhu Nath Ray; Partha Talukdar
err分享
err收藏
RE-GCN: Relation Enhanced Graph Convolutional Network for Entity Alignment in Heterogeneous Knowledge Graphs
errLecture Notes in Computer Science
IF0
err2020-09-22
err0
PREAI
errJinzhu Yang; Wei Zhou; Lingwei Wei; Junyu Lin; Jizhong Han; Songlin Hu
err分享
err收藏
Learning Sequence Encoders for Temporal Knowledge Graph Completion
errProceedings of the 2018 Conference on Empirical Methods in Natural Language Processing
IF0
err2018-01-01
err0
errOAAI
errAlberto García-Durán; Sebastijan Dumančić; Mathias Niepert
err分享
err收藏
Multi-Granularity History and Entity Similarity Learning for Temporal Knowledge Graph Reasoning
errProceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
IF0
err2024-01-01
err0
PREAI
errMingcong,Shi; Zhu,Chunjiang; Zhang,Detian; Wen,Shiting; Qing,Li
err分享
err收藏
Temporal Knowledge Graph Forecasting Without Knowledge Using In-Context Learning
errProceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
IF0
err2023-01-01
err0
errOAAI
errDong-Ho Lee; Kian Ahrabian; Woojeong Jin; Fred Morstatter; Jay Pujara
err分享
err收藏
学者 查看更多内容