1
Return

MPRG:A unified framework for knowledge graph reasoning via pattern-aware relation graph

delete2025-12-23
delete0
PRE
AI
谢志文 cover
谢志文 (Zhiwen Xie)
M
Meiling Wang
胡泊 cover
胡泊 (Po Hu)
DOI：10.1016/j.ipm.2025.104570delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• We propose a novel knowledge graph reasoning model MPRG that explicitly captures complex relational patterns. • The proposed model constructs a pattern-aware relation graph by mining logical rules, enabling shared relational semantics and structural priors across knowledge graphs. • The proposed model adopts a relation-entity collaborative reasoning framework, in which enriched relation embeddings guide entity-level inference for improved generalization to unseen entities and relations. • Experimental results show that MPRG consistently achieves state-of-the-art performance, proving its effectiveness and adaptability across various reasoning scenarios.

Journal

I
Information Processing and Management
IF:
6.9
Papers:
5.2K
Citations:
1.4W

Organization

C
Central China Normal University
Scholars:
1.1W
Papers: 8.1K
Citations: 1.1W
Cited Papers

Cited Papers

Citing Papers

Citing Papers