Identifiability of Cross-Domain Recommendation via Causal Subspace Disentanglement.
Jing DuZesheng YeBin GuoZhiwen YuLina YaoPublished in: SIGIR (2024)
Keyphrases
- cross domain
- causal models
- causal discovery
- causal bayesian networks
- causal relationships
- recommender systems
- transfer learning
- domain adaptation
- collaborative filtering
- sentiment classification
- knowledge transfer
- low dimensional
- principal component analysis
- multiple domains
- text categorization
- high dimensional data
- dimensionality reduction
- high dimensional
- directed acyclic graph
- cross domain learning
- target domain
- feature space
- neural network
- observational data
- e government
- linear subspace
- learning tasks
- text classification