Cold start system and method for dialog strategy optimization

Abstract

The invention relates to a dialogue strategy-optimized cold start system and a method. The system comprises a user input module, a dialogue state tracking module, a teacher decision-making module, a student decision-making module, an action selection module, an output module, a strategy training module and a reward function module. The action selection module randomly selects one final reply action from all reply actions generated by the teacher decision-making module and the student decision-making module. The output module converts the final reply action into a more natural expression and displays the more natural expression to a user. The strategy training module stores the dialogue experience (transition) in an empirical pool, samples a fixed number of experiences, and updates network parameters according to a depth Q network (DQN) algorithm. The reward function module calculates the reward of the dialogue at each round of the dialogue, and outputs the reward to the strategy training module. According to the invention, the performance of the dialogue strategy during the strengthened learning on-line training initial stage can be remarkably improved. The learning speed of the dialogue strategy is increased, and the number of dialogues used for achieving certain performances is reduced.

Description

technical field [0001] The invention relates to a technology in the field of intelligent man-machine dialogue, in particular to a cold start system and method for dialogue strategy optimization. Background technique [0002] The intelligent human-computer dialogue system is an intelligent system that can communicate with users. Among them, the dialog strategy is the module in the whole system that decides how to reply to the user. The earliest design method of dialogue strategy is that designers design different logic rules according to different user inputs. The disadvantage of this method is that the dialogue strategy cannot be continuously optimized with the user's feedback to enhance the adaptive ability to the user and the environment. [0003] In recent years, deep reinforcement learning methods have gradually been used in the optimization of dialogue policies. In this method, the dialogue strategy is represented by a neural network, and the reward signal (reward) i...

AI Technical Summary

Problems solved by technology

[0004] The present invention aims at the defect that the system performance is very poor at the initial stage of training in the prior art, and a large amount of dialogue data is required for training to improve the performance, and proposes a cold start system and method for dialogue strategy optimization, which can significantly improve dialogue strategy in reinforcement learning. Performance in the early stages of online training; improve the learning speed of the dialogue policy, that is, reduce the number of dialogues it uses to achieve a certain performance

Images