Finite-to-Infinite N-Best POMDP for Spoken Dialogue Management.
Guohua WuCaixia YuanBing LengXiaojie WangPublished in: CCL (2015)
Keyphrases
- dialogue management
- spoken dialogue
- partially observable markov decision process
- language understanding
- decision theoretic
- state space
- partially observable
- belief state
- dialogue system
- semi autonomous
- reinforcement learning
- spoken dialogue systems
- partially observable markov decision processes
- multi agent
- natural language understanding
- dynamical systems
- finite state
- markov decision processes
- markov chain
- natural language
- language processing
- heuristic search
- general knowledge
- optimal policy
- planning under uncertainty