2005 EMNLP EMNLP 2005

Learning Mixed Initiative Dialog Strategies By Using Reinforcement Learning On Both Conversants