Method and device for generating reply voice

Abstract

The invention discloses a method and device for generating reply voice and relates to the technical field of natural language processing. The method comprises the following steps that voice to be replied is received; whether text content corresponding to the voice to be replied is a target specific word or not is judged based on a target automatic voice recognition module, the target automatic voice recognition module is a function module pre-integrated in a soft switch subsystem, and the target specific word is any one of a plurality of preset positive words and a plurality of preset negative words; if yes, reply voice corresponding to the voice to be replied is generated according to the target specific word, a target natural language understanding module and a target character-to-voice module, and the target natural language understanding module and the target character-to-voice module are functional modules integrated in a soft switch subsystem in advance; and if not, the voice is sent to be replied to the intelligent voice subsystem so that the intelligent voice subsystem generates a reply voice corresponding to the voice to be replied.

Description

technical field [0001] The present application relates to the technical field of natural language processing, and in particular to a method and device for generating reply speech. Background technique [0002] With the continuous development of artificial intelligence technology, intelligent voice technology has also developed rapidly. The softswitch intelligent voice interaction system using intelligent voice technology can replace the customer service of mobile communication operators to conduct multiple rounds of conversations with customers. Among them, the softswitch intelligent voice interaction system includes softswitch subsystem and intelligent voice subsystem. Mobile communication operators receive After receiving the call voice from the customer, send the call voice to the soft switch subsystem, so that the soft switch subsystem sends the call voice to the intelligent voice subsystem. The intelligent voice subsystem is based on the automatic voice recognition modu...

AI Technical Summary

Problems solved by technology

[0003] In order to ensure the accuracy of recognizing complex call speech, the automatic speech recognition module in the intelligent voice subsystem usually performs speech recognition on the call speech based on the DNN-HMM (Deep Neural Network-Hidden Markov) model, thereby converting the call speech into text Content, however, based on the DNN-HMM model to perform speech recognition on call speech, it takes a long time; and, usually, the softswitch subsystem and the intelligent voice subsystem are deployed in different servers, so that the softswitch subsystem Communication with the intelligent voice subsystem requires network communication, therefore, there is a network transmission delay when the softswitch subsystem communicates with the intelligent voice subsystem

Therefore, it takes a long time for the existing softswitch intelligent voice interaction system to generate a reply voice, which leads to poor user experience for customers

Images