Speech processing method and device

Abstract

The invention discloses a speech processing method, which comprises the steps of classifying standard service texts according to preset service types, and determining the service type of each standard service text; recognizing the acquired speech, and converting the acquired speech into a text; determining basic language unit of the converted text and a language function category of each basic language unit; determining a weight vector matrix formed by weight vectors corresponding to the basic language units by adopting a vector space model; determining the service type of the basic language unit corresponding to each weight vector in the weight vector matrix according to the standard service texts and the service types thereof and the weight vector matrix; and enabling the basic language units to form an output text according to the service type and the language function category of each basic language unit, and performing post-processing on the output text. The invention further discloses a speech processing device.

Description

technical field [0001] The invention relates to speech processing technology, in particular to a speech processing method and device. Background technique [0002] The call center is also called the customer service center. The main function of the call center is to have trained customer service personnel answer and deal with various problems of incoming customers through hotlines and consultation calls, such as inquiries, complaints, suggestions, etc.; [0003] The existing customer service mode is limited by the personal ability of the customer service personnel, and the service quality is inconsistent due to the customer service personnel's intonation difference, language difference speed, terminology standardization, business understanding, business familiarity, business explanation accuracy, etc. There are even some mistakes; therefore, the service received by customers is a non-standardized and non-standardized service, which will be affected by differences in customer...

AI Technical Summary

Problems solved by technology

[0003] The existing customer service mode is limited by the personal ability of the customer service personnel, and the service quality is inconsistent due to the customer service personnel's intonation difference, language difference speed, terminology standardization, business understanding, business familiarity, business explanation accuracy, etc. There are even some mistakes; therefore, the service received by customers is a non-standardized and non-standardized service, which will be affected by differences in customer service personnel;

Images