Reinforcement Learning to Optimize Long-term User Engagement in Recommender Systems.
Lixin ZouLong XiaZhuoye DingJiaxing SongWeidong LiuDawei YinPublished in: KDD (2019)
Keyphrases
- user engagement
- recommender systems
- long term
- reinforcement learning
- user experience
- short term
- collaborative filtering
- matrix factorization
- user preferences
- cold start problem
- function approximation
- markov decision processes
- multi agent
- temporal difference
- user profiling
- reinforcement learning algorithms
- model free
- information retrieval
- learning process
- transfer learning
- user satisfaction
- optimal policy
- machine learning
- recommendation quality
- information overload
- user modeling
- state space
- trust aware
- social media
- dynamic programming
- implicit feedback
- end users
- data sparsity
- recommendation systems
- user modelling
- information filtering
- user feedback
- user model