HyperBandit: Contextual Bandit with Hypernewtork for Time-Varying User Preferences in Streaming Recommendation.
Chenglei ShenXiao ZhangWei WeiJun XuPublished in: CoRR (2023)
Keyphrases
- user preferences
- contextual bandit
- upper confidence bound
- news recommendation
- recommender systems
- collaborative filtering
- user behavior
- recommendation systems
- data streams
- user behaviour
- collaborative recommendation
- user profiles
- user feedback
- user specific
- preference model
- preference models
- decision making
- news items
- news articles
- skyline queries
- data sparsity
- cold start
- online news
- active learning