UP-NRPA: User Portrait based Nested Rollout Policy Adaptation for Planning with Large Language Models in Goal-oriented Dialogue Systems

The paper introduces UP-NRPA, a User Portrait based Nested Rollout Policy Adaptation framework designed for goal-oriented dialogue systems utilizing Large Language Models. This approach allows for real-time adaptation of dialogue strategies based on user feedback and characteristics, achieving a 100% success rate in various dialogue tasks and a 56.41% increase in negotiation task performance without traditional offline reinforcement learning. This innovation is significant for practitioners as it enables more responsive and personalized dialogue systems that can effectively cater to diverse user needs.