SaviorRec uses a behavior-aware multimodal encoder and residual quantized semantic IDs to improve cold-start CTR prediction, reporting offline and online gains on Taobao.
To improve recommendation performance for cold-start and long-tail items, re- cent studies focus on leveraging item multimodal features to model users’ interests
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
SaviorRec: Semantic-Behavior Alignment for Cold-Start Recommendation
SaviorRec uses a behavior-aware multimodal encoder and residual quantized semantic IDs to improve cold-start CTR prediction, reporting offline and online gains on Taobao.