RecMind aligns LLM text embeddings with LightGCN collaborative embeddings via contrastive learning and a learned gate, achieving the best reported scores on all 8 ranking metrics across two datasets.
Graph convolutional neural networks for web-scale recommender systems,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
RecMind: LLM-Enhanced Graph Neural Networks for Personalized Consumer Recommendations
RecMind aligns LLM text embeddings with LightGCN collaborative embeddings via contrastive learning and a learned gate, achieving the best reported scores on all 8 ranking metrics across two datasets.