.png){kind=icon}
.png){kind=icon}
Return
Knowledge graph contrastive learning for recommendation via knowledge-aware reasoning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.eswa.2025.130900.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• We propose KGCR, a unified knowledge-aware reasoning framework that integrates relation-aware aggregation with cross-view contrastive learning. • A perception score is designed and reused to guide triple weighting and dual-view graph enhancement, enabling semantics-consistent augmentation. • Experiments and quantitative path analysis show that KGCR significantly improves recommendation performance, robustness, and explainability.
Journal
IF:
7.5
Papers:
2.9W
Citations:
10.2W