Guided Dialogue Policy Learning without Adversarial Learning in the Loop

32271

Knowledge Triples

Source Entity	Relationship	Target Entity
Adversarial Learning Method	learns	Reward Function Method
Dialogue Policy Material	measuredBy	Task Success Rate Metric
Local Optima OtherEntity	affects	Adversarial Learning Method
Mode Collapse OtherEntity	affects	Adversarial Learning Method
Policy Gradient Method	isA	Reinforcement Learning Method
REINFORCE Method	isA	Policy Gradient Method
User Simulator Material	provides	Reward Function Method