Reinforcement Learning-based Recommender Systems with Large Language Models for State Reward and Action Modeling.
Jie WangAlexandros KaratzoglouIoannis ArapakisJoemon M. JosePublished in: SIGIR (2024)
Keyphrases
- language model
- reinforcement learning
- recommender systems
- state space
- language modeling
- action space
- state action
- agent learns
- speech recognition
- probabilistic model
- n gram
- document retrieval
- language modelling
- action selection
- information retrieval
- retrieval model
- query expansion
- statistical language models
- reward shaping
- reward function
- smoothing methods
- statistical language modeling
- context sensitive
- markov decision processes
- test collection
- reinforcement learning algorithms
- document ranking
- discounted reward
- ad hoc information retrieval
- average reward
- reward signal
- agent receives
- language model for information retrieval
- translation model
- relevance model
- query terms
- optimal policy
- user profiles
- image retrieval
- co occurrence
- collaborative filtering