Diff-MSR: A Diffusion Model Enhanced Paradigm for Cold-Start Multi-Scenario Recommendation.
Yuhao WangZiru LiuYichao WangXiangyu ZhaoBo ChenHuifeng GuoRuiming TangPublished in: WSDM (2024)
Keyphrases
- cold start
- diffusion model
- recommender systems
- cold start problem
- collaborative filtering
- data sparsity
- personalized recommendation
- tag recommendation
- anisotropic diffusion
- implicit feedback
- user preferences
- information diffusion
- recommendation algorithms
- diffusion process
- data sparseness
- user ratings
- matrix factorization
- search engine
- recommendation systems
- user interests
- user feedback
- transfer learning
- active learning
- social networks