.png){kind=icon}
.png){kind=icon}
Return
Multiple minds are better than one: Enhancing temporal knowledge graph forecasting with mixture of diverse graph experts
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ipm.2026.104847.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• FMOGE first introduces dual-system theory for semantic evidence extraction via fast–slow thinking in TKGF. • FMOGE is a flexible graph-expert mixture that fuses structural and semantic information to enhance TKGF. • Three experts model event dynamics, context, and logic, unified by a hierarchical merger for reasoning. • FMOGE significantly outperforms 13 state-of-the-art baselines.
Keywords:
Temporal Knowledge Graph Forecasting
Semantic Evidence Extraction
Graph-Expert Mixture
Dual-System Theory
Event Dynamics
Journal
I
IF:
6.9
Papers:
293
Citations:
0