Multi-level dialogue analysis method for intelligent voice dialogue system

Abstract

The invention belongs to the field of artificial intelligence, and particularly relates to a multi-level dialogue analysis method for an intelligent voice dialogue system. The method aims at solving the problem that an existing Chinese spoken dialogue tagging system is imperfect, and improves the naturalness and accuracy of dialogue understanding. The method comprises the steps of conducting voicerecognition of input voice to obtain corresponding text; combining the voice with text information to conduct a rhythmic-level analysis; conducting a syntactic-level analysis on the text; conductinga semantic-level analysis on the text; conducting a pragmatic-level analysis on the text. The method is utilized at the four levels of rhythm, syntax, semantics and pragmatics, dimensional sets at allthe levels are determined according to language unit levels from a word level to a passage level, analyses at all the levels are combined so that the intelligent dialogue system can be used for understanding the intention of others' conversations better, and the user experience is improved.

Description

technical field [0001] The application belongs to the field of artificial intelligence, relates to an intelligent dialogue agent system, and in particular relates to a multi-level dialogue analysis method for an intelligent voice dialogue system. Background technique [0002] Dialogue system is an important branch of the field of artificial intelligence, and has a wide range of applications in voice customer service agents, intelligent robots and other fields. Its main technologies include recognizing speech as text, performing natural language understanding on text, finding answers to understood questions, and outputting answers by voice. The natural language understanding part of the dialogue system is the most important. [0003] In recent years, foreign countries regard the automatic recognition of "dialogue acts" as the key to the natural language understanding system of oral conversation, and the most basic task is to establish a classification system for conversation...

AI Technical Summary

Problems solved by technology

[0006] Aiming at the problem that the existing spoken Chinese dialogue annotation system is not perfect, the present invention provides a multi-level dialogue analysis method for intelligent speech dialogue system, which includes interrelated dimensional information of prosody, syntax, semantics and pragmatics. The analysis of the interaction between language units at different levels leads to in-depth dialogue analysis and improves the naturalness and accuracy of dialogue understanding

Images