A Mixture-of-Expert Approach to RL-based Dialogue Management.
Yinlam ChowAza TulepbergenovOfir NachumDhawal GuptaMoonkyung RyuMohammad GhavamzadehCraig BoutilierPublished in: ICLR (2023)
Keyphrases
- dialogue management
- dialogue system
- reinforcement learning
- spoken dialogue systems
- learning agent
- natural language generation
- language understanding
- virtual humans
- multi agent
- model free
- markov decision processes
- optimal policy
- learning process
- reinforcement learning algorithms
- domain experts
- state space
- multi domain
- human machine interaction
- knowledge acquisition
- partially observable markov decision process