Controllable dialogue management extension method based on fusion rule information

Abstract

The invention belongs to the technical field of man-machine dialogues, particularly relates to a controllability dialogue management extension method capable of carrying out rule information fusion, and aims to solve the problems of high cost and low efficiency when a data-driven dialogue system carries out extension through a way of interactive environment reconstruction. The method comprises thefollowing steps that: S1: on the basis of interaction data, determining new user intentions which need to be extended, and carrying out extension on an original language comprehension module; S2: onthe basis of the new user intentions, constructing a new dialogue rule corresponding to the new user intentions; S3: on the basis of the interaction data, the dialogue strategy of an original dialoguemanagement module and a new dialogue rule, constructing a constraint which needs to be met by a new dialogue management module mapping space; and S4: on the basis of the constraint which needs to bemet by the new dialogue management module mapping space obtained in S3, carrying out extension on the original dialogue management module, and generating a new dialogue management module. By use of the method, the data-driven dialogue system can be subjected to quick extension and efficient iteration according to user feedback.

Description

technical field [0001] The invention belongs to the technical field of man-machine dialogue, and in particular relates to a controllable dialogue management extension method for fusing rule information. Background technique [0002] A task-based dialogue system is a human-computer interaction system that assists users to complete tasks in a specific field (restaurants, hotels, or airline tickets) through natural language interaction. Generally speaking, a task-based dialogue system needs to have the following four basic functions: language understanding, dialogue state tracking, dialogue strategy and dialogue generation. Among them, the dialog state tracking module and the dialog policy module are called the dialog management module together, which is the core part of the whole system. [0003] Because the task-based dialogue system can help consumers achieve their goals in a more friendly way. Therefore, people have invested a lot of energy to study how to design a dialog...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the above problems in the prior art, that is, in order to solve the problem of high cost and low efficiency when expanding the data-driven dialogue system, the present invention proposes a controllable dialogue management extension method of fusing rule information, including the following step:

Images