GCIB denoises auxiliary behavior graphs via graph information bottleneck and enriches target embeddings through cross-behavior graph contrastive learning for improved multi-behavior recommendation.
Less is more: Information bottleneck denoised multime- dia recommendation.arXiv preprint arXiv:2501.12175,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
GCIB: Graph Contrastive Information Bottleneck for Multi-Behavior Recommendation
GCIB denoises auxiliary behavior graphs via graph information bottleneck and enriches target embeddings through cross-behavior graph contrastive learning for improved multi-behavior recommendation.