Causal Distillation for Alleviating Performance Heterogeneity in Recommender Systems.
Shengyu ZhangZiqi JiangJiangchao YaoFuli FengKun KuangZhou ZhaoShuo LiHongxia YangTat-Seng ChuaFei WuPublished in: CoRR (2024)
Keyphrases
- recommender systems
- collaborative filtering
- causal relationships
- bayesian networks
- user profiling
- matrix factorization
- user modeling
- causal inference
- information overload
- causal theories
- causal relations
- causal networks
- causal knowledge
- information filtering
- trust aware
- user preferences
- user profiles
- causal models
- product recommendation
- item based collaborative filtering
- neural network
- causal reasoning
- personalized recommendation
- recommendation systems
- cold start
- semantic heterogeneity
- user modelling
- data sparsity
- causal independence
- directed acyclic graph
- user model