CN114818649A - Business consultation processing method and device based on intelligent voice interaction technology - Google Patents

Abstract

The embodiment of the invention discloses a service consultation processing method and a device based on an intelligent voice interaction technology, which can be used in the financial field or other technical fields, and the method comprises the following steps: acquiring business consultation voice of a client; identifying a business consultation text corresponding to the business consultation voice; similarity matching is carried out on the service consultation text and a preset service consultation template, and the service consultation template with the highest similarity to the service consultation text is determined; and if the similarity corresponding to the service consultation template with the highest similarity to the service consultation text is greater than or equal to a preset threshold value, sending the service consultation response information corresponding to the service consultation template with the highest similarity to the service consultation text to the client. The invention realizes that simple and repeated consultation services are automatically completed by the intelligent voice system, improves the service quality and efficiency to a certain extent, reduces the waiting time of users and reduces the working intensity of manual customer service response.

Description

Business consultation processing method and device based on intelligent voice interaction technology

technical field

The present invention relates to the technical field of artificial intelligence, in particular, to a business consultation processing method and device based on intelligent voice interaction technology.

Background technique

With the advent of the era of artificial intelligence, intelligent speech recognition, image recognition, natural language processing and other related technologies are becoming more and more mature, and there is great potential and value in applying them to the traditional financial service industry. Under the premise of the continuous development of information technology, customer service personnel in banks and other related industries need to accept more business knowledge to match more mature new technologies; in addition, due to the influence of human and material resources, adding customer service personnel is not the best choice. . In this situation, the work types of customer service personnel are single and repetitive, and the low access rate of incoming calls will lead to poor service experience for users and lead to a large number of complaints. The current customer service system in the financial industry mainly includes call centers and human agents (human customer service).

The call center is also the incoming and outgoing end of the phone. Generally, customers use mobile devices such as phones and mobile phones to connect to the call center for business consultation and related processing. The intelligent call platform of the telephone customer service system is relayed and interactive voice response (IVR). ) platform to perform task transfer, and transfer to manual customer service for operation when the customer needs a manual response. In the financial industry, such as banks, manual customer service is generally used to provide real-time feedback to customers' business inquiries and operations, and to provide customers with services such as business consulting, auxiliary transactions, troubleshooting, and outbound marketing. Complaints, timely submission of work contact orders and transfer to the background for processing, tracking the processing situation, and responding to customers in a timely manner.

The traditional customer service system has a high average daily call repetition rate due to the single business. The traditional method of manual customer service connection has higher requirements for customer service personnel. During the busy period of manual customer service, customers will wait for a long time, the answering rate is low, and customer complaints gradually increase. The increase will seriously affect the service experience.

SUMMARY OF THE INVENTION

In order to solve at least one technical problem in the above-mentioned background technology, the present invention proposes a business consultation processing method and device based on intelligent voice interaction technology.

In order to achieve the above object, according to an aspect of the present invention, a method for processing business consultation based on intelligent voice interaction technology is provided, the method comprising:

Obtain customer's business consultation voice;

Identifying the business consultation text corresponding to the business consultation voice;

Matching the business consultation text with a preset business consultation template for similarity, and determining the business consultation template with the highest similarity with the business consultation text;

If the similarity corresponding to the business consultation template with the highest similarity to the business consultation text is greater than or equal to a preset threshold, send the business consultation answer information corresponding to the business consultation template with the highest similarity to the business consultation text to said client.

Optionally, identifying the business consultation text corresponding to the business consultation voice specifically includes:

Framing processing is performed on the business consultation voice, and the waveform of each frame is converted into a sound feature;

Inputting the sound feature into a preset acoustic model to obtain phoneme information corresponding to each of the sound features output by the acoustic model;

According to the corresponding relationship between the factor information and the word, determine the word or word corresponding to each of the phoneme information;

The business consultation text is generated according to the corresponding character or word of each of the phoneme information.

Optionally, performing similarity matching between the business consultation text and a preset business consultation template, and determining a business consultation template with the highest similarity with the business consultation text, specifically including:

determining the business type corresponding to the business consultation text;

The business consultation texts are respectively matched with all the business consultation templates corresponding to the business types to determine the business consultation template with the highest similarity with the business consultation text.

Optionally, the determining the business type corresponding to the business consultation text specifically includes:

Use long-short-term memory network to perform feature extraction on the business consulting text to obtain text features;

According to the text feature, the convolutional neural network is used to classify and identify the text, so as to obtain the business type corresponding to the business consultation text.

Optionally, the business consultation answer information includes: business consultation answer text;

The sending of the business consultation response information corresponding to the business consultation template with the highest similarity to the business consultation text to the client specifically includes:

Convert the business consultation answer text corresponding to the business consultation template with the highest similarity to the business consultation text into an answer voice, and send the answer voice to the call center system, so that the call center system sends the answer voice Play to the client.

Optionally, converting the business consultation answer text corresponding to the business consultation template with the highest similarity to the business consultation text into an answer voice, specifically including:

Use natural language processing method to extract text language features on the business consultation answer text;

The extracted text language features are clustered using decision trees, and spectral coefficients and audio flags are obtained according to the clustering results and the hidden Markov model;

According to the obtained spectral coefficients and audio flag bits, the maximum likelihood parameter generation algorithm is used to sample the spectral coefficients;

The spectral coefficient sampling result is input into the vocoder to generate the speech waveform, and the answer speech is obtained.

Optionally, the business consultation processing method based on the intelligent voice interaction technology further includes:

If the highest similarity is less than the threshold, send a no-match answer instruction to the call center system, so that the call center system connects the customer's call to a manual customer service terminal according to the no-match answer instruction.

In order to achieve the above object, according to another aspect of the present invention, a service consultation processing device based on intelligent voice interaction technology is provided, the device comprising:

The business consultation voice acquisition unit is used to obtain the customer's business consultation voice;

a voice recognition unit, configured to identify the business consultation text corresponding to the business consultation voice;

a similarity matching unit, configured to perform similarity matching between the business consultation text and a preset business consultation template, and determine the business consultation template with the highest similarity with the business consultation text;

A business consultation response information sending unit, configured to send the business consultation template with the highest similarity to the business consultation text if the similarity corresponding to the business consultation template with the highest similarity is greater than or equal to a preset threshold The business consultation response information corresponding to the consultation template is sent to the customer.

In order to achieve the above object, according to another aspect of the present invention, a computer device is also provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the computer The program implements the steps of the above-mentioned business consultation processing method based on the intelligent voice interaction technology.

In order to achieve the above object, according to another aspect of the present invention, there is also provided a computer-readable storage medium on which a computer program/instruction is stored, and when the computer program/instruction is executed by a processor, the above-mentioned intelligent voice-based interaction technology is realized steps of the business consulting approach.

In order to achieve the above object, according to another aspect of the present invention, a computer program product is also provided, comprising a computer program/instruction, when the computer program/instruction is executed by a processor, the above-mentioned business consultation processing method based on intelligent voice interaction technology is implemented A step of.

The beneficial effects of the present invention are:

In the embodiment of the present invention, the business consultation text corresponding to the business consultation voice input by the customer is first identified, and then the similarity between the business consultation text and the preset business consultation template is matched to determine the business with the highest similarity with the business consultation text. Consulting template, and finally if the similarity corresponding to the business consulting template with the highest similarity to the business consulting text is greater than or equal to the preset threshold, then the business consulting template corresponding to the business consulting template with the highest similarity to the business consulting text The consultation answer information is sent to the customer, and the present invention uses the intelligent voice recognition technology to perform intelligent business distribution, and through the establishment of a business consultation template and corresponding business consultation answer information, the simple and repeated consultation business is automatically completed by the intelligent voice system. , to a certain extent, improve the service quality and efficiency, reduce the user's waiting time, and reduce the work intensity of manual customer service response.

Description of drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are For some embodiments of the present invention, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without creative effort. In the attached image:

1 is a flowchart of a business consultation processing method based on an intelligent voice interaction technology according to an embodiment of the present invention;

2 is a flowchart of identifying a business consultation text corresponding to a business consultation voice according to an embodiment of the present invention;

3 is a flowchart of similarity matching according to an embodiment of the present invention;

4 is a flow chart of determining a business type corresponding to a business consultation text according to an embodiment of the present invention;

Fig. 5 is the flow chart that the embodiment of the present invention converts business consultation answer text into answer voice;

6 is a logical architecture diagram of an intelligent voice system according to an embodiment of the present invention;

7 is an overall flow chart of a business consultation processing method based on an intelligent voice interaction technology according to an embodiment of the present invention;

8 is a structural block diagram of a service consultation processing device based on an intelligent voice interaction technology according to an embodiment of the present invention;

FIG. 9 is a schematic diagram of a computer device according to an embodiment of the present invention.

Detailed ways

In order to make those skilled in the art better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only Embodiments are part of the present invention, but not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

It should be noted that the terms "comprising" and "having" in the description and claims of the present invention and the above-mentioned drawings, as well as any variations thereof, are intended to cover non-exclusive inclusion, for example, including a series of steps or units The processes, methods, systems, products or devices are not necessarily limited to those steps or units expressly listed, but may include other steps or units not expressly listed or inherent to such processes, methods, products or devices.

It should be noted that the embodiments of the present invention and the features of the embodiments may be combined with each other under the condition of no conflict. The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

It should be noted that the acquisition, storage, use, and processing of data in the technical solution of the present application all comply with the relevant provisions of national laws and regulations.

The purpose of the present invention is to establish a business consultation processing method based on intelligent voice interaction technology. The present invention adopts intelligent voice interaction, deep learning, big data analysis and other technologies to add intelligent voice systems to some common scenarios to realize intelligent function diversion. , realize intelligent voice response interaction for business inquiries with high inquiries such as card business processing, balance inquiry, card loss reporting, mobile phone number modification, etc. Waiting rate, improve business processing efficiency and service satisfaction.

The implementation main body of the business consultation processing method based on the intelligent voice interaction technology in the following embodiments of the present invention is the intelligent voice system in FIG.

Fig. 1 is a flow chart of a business consultation processing method based on intelligent voice interaction technology according to an embodiment of the present invention. As shown in Fig. 1, in an embodiment of the present invention, the business consultation processing method based on intelligent voice interaction technology of the present invention includes steps S101 to step S104.

Step S101, acquiring the customer's service consultation voice.

As shown in FIG. 6 , in an embodiment of the present invention, the customer specifically inputs a service consultation voice in the call center system (call center telephone voice platform). In one embodiment of the present invention, the customer accesses the call center system (call center phone voice platform) by telephone or network voice, and enters service consultation voice according to the instructions in the dialogue with the automatic customer service robot of the call center system.

Step S102, identifying the business consultation text corresponding to the business consultation voice.

As shown in FIG. 6 , the present invention converts the customer's business consultation voice into text information through a preset intelligent voice recognition module. In an embodiment of the present invention, the customer conducts business consultation or business processing through a mobile communication device, directly enters the intelligent voice system in the call center system, and "speaks" his own questions or needs under the guidance of the prompt tone of the intelligent voice self-service system . The IVR platform in the call center system sends the user's voice to the intelligent voice recognition module, and transcribes the business consultation voice input by the customer into text through intelligent voice recognition.

In step S103, the similarity between the business consultation text and the preset business consultation template is matched, and the business consultation template with the highest similarity with the business consultation text is determined.

As shown in FIG. 6 , the present invention performs semantic analysis through a preset knowledge retrieval core module, performs matching in the knowledge base, and finds answers to relevant questions of the customer's business consultation. In one embodiment of the present invention, after transcribing the voice input by the customer into text, the present invention invokes the knowledge retrieval core module to perform intelligent analysis on the transcribed text in combination with the information and rules on the knowledge management optimization server, and after obtaining the analysis result, Return the analysis results to the IVR platform. Complete the function of "listening" and "thinking" of intelligent customer service. The core module of knowledge retrieval conducts a detailed analysis of the text transcribed by the user's voice to understand the user's needs. The knowledge retrieval core module judges and finds the relevant question answers of the customer's business consultation, and can use the intelligent customer service speech synthesis server to be responsible for sending the question answers to the client. If the business of complex scenarios cannot be processed by the intelligent voice system, it can be transferred to a manual agent for processing.

Step S104, if the similarity corresponding to the business consultation template with the highest similarity to the business consultation text is greater than or equal to a preset threshold, then the business consultation template corresponding to the business consultation template with the highest similarity to the business consultation text A reply message is sent to the client.

As shown in FIG. 6 , the present invention performs speech synthesis on text answers through a preset intelligent speech synthesis module, and converts the answers to questions to be played to customers through the IVR in the telephone voice platform of the call center. In one embodiment of the present invention, the intelligent speech synthesis module is responsible for synthesizing the content of the knowledge base, the IVR server invokes the intelligent speech synthesis module to synthesize the text of the knowledge base content into speech, and then the IVR server plays the synthesized speech to the user, through Answer or solve the user's problem by voice interaction, and complete the "speaking" function of intelligent customer service. According to the user's needs, the IVR obtains the answer that the user needs from the knowledge base and returns it to the customer.

In an embodiment of the present invention, the business consultation processing method based on the intelligent voice interaction technology of the present invention further includes:

If the highest similarity is less than the threshold, send a no-match answer instruction to the call center system, so that the call center system connects the customer's call to a manual customer service terminal according to the no-match answer instruction.

As shown in FIG. 7 , in the embodiment of the present invention, the business consultation text corresponding to the business consultation voice input by the customer is first identified, and then the similarity between the business consultation text and the preset business consultation template is matched, and the similarity between the business consultation text and the business consultation template is determined. The business consulting template with the highest text similarity, and finally if the similarity corresponding to the business consulting template with the highest similarity with the business consulting text is greater than or equal to a preset threshold, then the business consulting template with the highest similarity with the business consulting text is determined. The business consultation answer information corresponding to the business consultation template is sent to the customer, and if the similarity corresponding to the business consultation template with the highest similarity to the business consultation text is less than a preset threshold, the call center system is notified to call the customer. Access to human customer service for processing.

It can be seen that the present invention adopts intelligent voice interaction, deep learning, big data analysis and other technologies to add intelligent voice systems in some common scenarios to realize intelligent function diversion, and realize intelligent voice response interaction for some common business consultations, which is helpful for To improve business processing efficiency and service satisfaction.

In an embodiment of the present invention, the present invention further adds an evaluation index of the intelligent voice system. After answering the questions for the customer, the customer service evaluation and the customer's response intention are analyzed to comprehensively judge the customer's satisfaction. The intelligent voice system is optimized through two evaluation systems, manual annotation and user feedback, and tags are classified for customers.

As shown in FIG. 2 , in an embodiment of the present invention, identifying the business consultation text corresponding to the business consultation voice in the above step S102 specifically includes steps S201 to S204 .

In step S201, the service consultation voice is divided into frames, and the waveform of each frame is converted into sound features.

In an embodiment of the present invention, before step S201 is performed, the present invention also removes the mute at the beginning and end of the service consultation voice input by the customer to reduce interference to subsequent steps.

In an embodiment of the present invention, this step specifically uses a moving window function to process the business consultation voice input by the customer into frames, and uses the Mel Cepstral Coefficient (MFCC) to transform each frame of waveform into a multi-dimensional vector containing sound information to obtain sound characteristics.

Step S202: Input the sound feature into a preset acoustic model to obtain phoneme information corresponding to each sound feature output by the acoustic model.

In an embodiment of the present invention, the input of the acoustic model is a feature vector, the output is phoneme information, and the acoustic model is obtained by training speech data.

Step S203, according to the corresponding relationship between the factor information and the word, determine the word or word corresponding to each of the phoneme information.

In the present invention, according to the correspondence between words or words and phonemes, Chinese is the correspondence between pinyin and Chinese characters, and English is the correspondence between phonetic symbols and words, to obtain the letters or Chinese corresponding to the factors.

Step S204, the business consultation text is generated according to the corresponding character or word of each of the phoneme information.

In the present invention, the present invention obtains the probability that a single word or words are related to each other by training a large amount of text information, and then extracts the feature based on the probability that the words are related to each other and the corresponding words or word pairs of the phoneme information. The audio data is output in text, and the voice input by the customer is converted into text.

As shown in FIG. 3, in an embodiment of the present invention, in the above step S103, the similarity between the business consultation text and the preset business consultation template is matched, and the business consultation template with the highest similarity with the business consultation text is determined. , specifically including step S301 and step S302.

Step S301: Determine the business type corresponding to the business consultation text.

Step S302: Match the business consultation text with all the business consultation templates corresponding to the business type respectively, and determine the business consultation template with the highest similarity with the business consultation text.

In an embodiment of the present invention, when performing similarity matching, the present invention can use BiMPM (bilateral multi-sided viewing angle matching model) to perform similarity matching. Firstly, the feature extraction of the two text objects to be matched is performed to obtain a weighted feature vector, and the similarity between the two weighted vectors is calculated by the similarity measure function cosine, and the matching degree of the two objects is output. In an embodiment of the present invention, for all calculated similarities, the present invention can use the softmax function to perform a normalization operation, determine the answer with the highest similarity, and return the most accurate answer corresponding to the customer's business consultation.

As shown in FIG. 4 , in an embodiment of the present invention, the determination of the service type corresponding to the service consultation text in the above step S301 specifically includes steps S401 and S402.

Step S401, using a long-short-term memory network to perform feature extraction on the business consultation text to obtain text features.

In one embodiment of the present invention, the present invention first uses LSTM (Long Short Term Memory Network) to extract features from text information, and the output of each time step includes not only the feature vector information of the current text but also the feature information of all previous texts.

Step S402, using a convolutional neural network to classify and identify the text according to the text feature, to obtain a business type corresponding to the business consulting text.

In one embodiment of the present invention, the present invention uses CNN (Convolutional Neural Network) to perform text classification and recognition on text features, preliminarily determine which business type the business belongs to, and perform matching in the knowledge base to find relevant question responses.

In an embodiment of the present invention, the business consultation answer information includes: business consultation answer text.

In an embodiment of the present invention, in the above step S104, the business consultation response information corresponding to the business consultation template with the highest similarity to the business consultation text is sent to the client, which specifically includes:

Convert the business consultation answer text corresponding to the business consultation template with the highest similarity to the business consultation text into an answer voice, and send the answer voice to the call center system, so that the call center system sends the answer voice Play to the client.

As shown in FIG. 5 , in an embodiment of the present invention, in the above steps, the business consultation answer text corresponding to the business consultation template with the highest similarity to the business consultation text is converted into an answer voice, which specifically includes steps S501 to S504 .

Step S501, using a natural language processing method to perform text language feature extraction on the business consultation answer text.

In an embodiment of the present invention, the conversion of text into speech is completed by a speech synthesis engine, and the process can be divided into two steps: front-end and back-end. The front-end is responsible for the grammatical analysis of the text, and uses NLP (Natural Language Processing) to extract the language features of the text. Through the specification of the dictionary and grammar, the formatted information at the grammar level is obtained and sent to the back-end.

In step S502, the extracted text language features are clustered using a decision tree, and spectral coefficients and audio flags are obtained according to the clustering results and the hidden Markov model.

Step S503, using the maximum likelihood parameter generation algorithm to sample the spectral coefficients according to the obtained spectral coefficients and the audio flag bit.

In an embodiment of the present invention, the back-end uses a decision tree to cluster the extracted text and language features based on the results of the front-end analysis, and performs a statistical parameter model HMM (Hidden Markov Model) calculation according to the clustering results. train. Use HMM to get spectral coefficients and audio flags. Then use MLPG (Maximum Likelihood Parameter Generation Algorithm) to sample the spectral coefficients.

Step S504, inputting the spectral coefficient sampling result into a vocoder to generate a speech waveform to obtain the answer speech.

In one embodiment of the present invention, the present invention inputs the spectral coefficient sampling result into the MLSA vocoder to generate the speech waveform. The prosodic and intonation are analyzed and processed to obtain basic information such as the length and intonation of the speech, and the most appropriate speech fragments are retrieved from this information, and the speech fragments are spliced and optimized to form the final played speech. In one embodiment of the present invention, the IVR platform plays the synthesized speech to the client.

In the present invention, since the answering ability of the intelligent speech system is limited by the richness of the knowledge base corpus, the optimal state of the intelligent speech system will be able to process and answer the questions with the highest accuracy. The present invention can also use the evaluation method of manual labeling for systematic evaluation, sample small data sets from the daily full data, try to ensure that the data distribution of the small data set is consistent with the full data, and use the accuracy rate, recall rate, rejection rate Rate and other indicators to evaluate system performance.

In the present invention, the present invention can also perform evaluation based on user feedback. After the phone call, the intelligent voice system conducts a comprehensive evaluation of the waiting market, customer satisfaction, and answer accuracy rate for the single business through big data statistical analysis by inquiring about customer satisfaction, and the key to customers' intentions through the business processing process Words for customer big data label classification, in order to better serve customers.

It can be seen from the above embodiments that the present invention adds an intelligent voice system to the customer service system, and uses intelligent voice recognition and synthesis technology to perform intelligent business diversion, which greatly reduces the waiting time of users and reduces the work stress of manual customer service responses. , for simple and repeated business, it is completed by the intelligent voice system, which improves the service quality and efficiency to a certain extent. The intelligent voice system optimizes and improves the traditional single artificial voice call. Customers do not need to press traditional buttons to handle the corresponding business. They only need to express their business needs and can conduct business processing through the intelligent voice system, which is convenient and efficient, and saves labor costs. Improve the overall service level of the call line, improve customer satisfaction, and continuously improve the user experience.

It should be noted that the steps shown in the flowcharts of the accompanying drawings may be executed in a computer system, such as a set of computer-executable instructions, and, although a logical sequence is shown in the flowcharts, in some cases, Steps shown or described may be performed in an order different from that herein.

Based on the same inventive concept, the embodiment of the present invention also provides a service consultation processing device based on the intelligent voice interaction technology, which can be used to implement the service consultation processing method based on the intelligent voice interaction technology described in the above embodiments, as follows. example. Since the principle of solving problems of the business consultation processing device based on the intelligent voice interaction technology is similar to the business consultation processing method based on the intelligent voice interaction technology, the embodiment of the business consultation processing device based on the intelligent voice interaction technology can refer to the The embodiments of the business consulting processing method will not be repeated here. As used below, the term "unit" or "module" may be a combination of software and/or hardware that implements a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, implementations in hardware, or a combination of software and hardware, are also possible and contemplated.

8 is a structural block diagram of a service consultation processing device based on intelligent voice interaction technology according to an embodiment of the present invention. As shown in FIG. 8 , in an embodiment of the present invention, the service consultation processing device based on intelligent voice interaction technology of the present invention includes:

A business consulting voice acquiring unit 1, used for acquiring the customer's business consulting voice;

A voice recognition unit 2, configured to identify the business consultation text corresponding to the business consultation voice;

Similarity matching unit 3, configured to perform similarity matching between the business consultation text and a preset business consultation template, and determine the business consultation template with the highest similarity with the business consultation text;

The business consultation answer information sending unit 4 is configured to, if the similarity corresponding to the business consultation template with the highest similarity with the business consultation text is greater than or equal to a preset threshold, then send the business consultation template with the highest similarity to the business consultation text. The business consultation answer information corresponding to the business consultation template is sent to the customer.

In an embodiment of the present invention, the service consultation processing device based on the intelligent voice interaction technology of the present invention further includes:

The no-matching answer instruction sending unit is configured to send a no-matching answering instruction to the call center system if the highest similarity is less than the threshold, so that the call center system sends the customer's call to the call center system according to the no-matching answering instruction. Access the manual customer service terminal.

In an embodiment of the present invention, the speech recognition unit 2 specifically includes:

a sound feature conversion module, which is used to perform frame division processing on the business consultation voice, and convert the waveform of each frame into a sound feature;

an input module, configured to input the sound feature into a preset acoustic model to obtain phoneme information corresponding to each of the sound features output by the acoustic model;

Correspondence search module, used for determining the corresponding word or word of each of the phoneme information according to the corresponding relationship between the factor information and the word;

A text generation module, configured to generate the business consultation text according to the corresponding word or word of each of the phoneme information.

In an embodiment of the present invention, the similarity matching unit 3 specifically includes:

a business type determination module, configured to determine the business type corresponding to the business consultation text;

The similarity calculation module is configured to perform similarity matching between the business consultation text and all the business consultation templates corresponding to the business type, and determine the business consultation template with the highest similarity with the business consultation text.

In an embodiment of the present invention, the service type determination module specifically includes:

A text feature extraction sub-module, used for extracting features from the business consulting text by using a long-short-term memory network to obtain text features;

The type determination sub-module is used for classifying and identifying the text by using the convolutional neural network according to the text feature, and obtaining the business type corresponding to the business consulting text.

In an embodiment of the present invention, the business consultation answer information includes: business consultation answer text.

In an embodiment of the present invention, the service consultation reply information sending unit 4 includes:

The voice conversion module is used to convert the business consultation answer text corresponding to the business consultation template with the highest similarity to the business consultation text into an answer voice, and send the answer voice to the call center system, so that the call center The system plays the answer voice to the customer.

In an embodiment of the present invention, the voice conversion module specifically includes:

A text language feature extraction sub-module, used for extracting text language features from the business consultation answer text by using a natural language processing method;

The clustering sub-module is used to cluster the extracted text language features using a decision tree, and obtain spectral coefficients and audio flags according to the clustering results and the hidden Markov model;

The spectral coefficient sampling sub-module is used to sample the spectral coefficients by using the maximum likelihood parameter generation algorithm according to the obtained spectral coefficients and audio flag bits;

The speech waveform generation sub-module is used for inputting the spectral coefficient sampling result into the vocoder to generate the speech waveform to obtain the answer speech.

In order to achieve the above object, according to another aspect of the present application, a computer device is also provided. As shown in FIG. 9 , the computer device includes a memory, a processor, a communication interface and a communication bus, and a computer program that can be run on the processor is stored in the memory, and the processor implements the above embodiments when executing the computer program steps in the method.

The processor may be a central processing unit (Central Processing Unit, CPU). The processor may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA) or other Chips such as programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above types of chips.

As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs, non-transitory computer-executable programs, and units, such as program units corresponding to the above method embodiments of the present invention. The processor executes various functional applications of the processor and works data processing by running the non-transitory software programs, instructions and modules stored in the memory, that is, to implement the methods in the above method embodiments.

The memory may include a storage program area and a storage data area, wherein the storage program area may store an operating system and an application program required by at least one function; the storage data area may store data created by the processor, and the like. Additionally, the memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one magnetic disk storage device, flash memory device, or other non-transitory solid state storage device. In some embodiments, the memory may optionally include memory located remotely from the processor, such remote memory being connectable to the processor via a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

The one or more units are stored in the memory, and when executed by the processor, perform the methods in the above-described embodiments.

The specific details of the above computer equipment can be understood by referring to the corresponding related descriptions and effects in the above embodiments, which will not be repeated here.

In order to achieve the above object, according to another aspect of the present application, a computer-readable storage medium is also provided, where the computer-readable storage medium stores a computer program, and when the computer program is executed in a computer processor, the above-mentioned based on Steps in a business consultation processing method of intelligent voice interaction technology. Those skilled in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. During execution, the processes of the embodiments of the above-mentioned methods may be included. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM), a random access memory (Random Access Memory, RAM), a flash memory (Flash Memory), a hard disk (Hard Disk Drive, Abbreviation: HDD) or solid-state drive (Solid-State Drive, SSD), etc.; the storage medium may also include a combination of the above-mentioned types of memories.

In order to achieve the above object, according to another aspect of the present application, a computer program product is also provided, including a computer program/instruction, when the computer program/instruction is executed by a processor, the above-mentioned business consultation processing method based on intelligent voice interaction technology is implemented A step of.

Obviously, those skilled in the art should understand that the above-mentioned modules or steps of the present invention can be implemented by a general-purpose computing device, which can be centralized on a single computing device, or distributed in a network composed of multiple computing devices Alternatively, they can be implemented with program codes executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or they can be integrated into The multiple modules or steps are fabricated into a single integrated circuit module. As such, the present invention is not limited to any particular combination of hardware and software.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims (11)

1. a business consulting processing method based on intelligent voice interaction technology, is characterized in that, comprising: Obtain customer's business consultation voice; Identifying the business consultation text corresponding to the business consultation voice; Matching the business consultation text with a preset business consultation template for similarity, and determining a business consultation template with the highest similarity with the business consultation text; If the similarity corresponding to the business consultation template with the highest similarity to the business consultation text is greater than or equal to a preset threshold, send the business consultation answer information corresponding to the business consultation template with the highest similarity to the business consultation text to said client. 2. The business consultation processing method based on intelligent voice interaction technology according to claim 1, wherein the identifying the business consultation text corresponding to the business consultation voice specifically comprises: Framing processing is performed on the business consultation voice, and the waveform of each frame is converted into a sound feature; Inputting the sound feature into a preset acoustic model to obtain phoneme information corresponding to each of the sound features output by the acoustic model; According to the corresponding relationship between the factor information and the word, determine the word or word corresponding to each of the phoneme information; The business consultation text is generated according to the corresponding character or word of each of the phoneme information. 3. The method for processing business consultation based on intelligent voice interaction technology according to claim 1, wherein the business consultation text is matched with a preset business consultation template for similarity, and it is determined that the business consultation Business consulting templates with the highest text similarity, including: determining the business type corresponding to the business consultation text; The business consultation texts are respectively matched with all the business consultation templates corresponding to the business types to determine the business consultation template with the highest similarity with the business consultation text. 4. The business consultation processing method based on intelligent voice interaction technology according to claim 3, wherein the determining the business type corresponding to the business consultation text specifically comprises: Use long-short-term memory network to perform feature extraction on the business consulting text to obtain text features; According to the text feature, the convolutional neural network is used to classify and identify the text, so as to obtain the business type corresponding to the business consultation text. 5. The business consultation processing method based on intelligent voice interaction technology according to claim 1, wherein the business consultation answer information comprises: business consultation answer text; The sending of the business consultation response information corresponding to the business consultation template with the highest similarity to the business consultation text to the client specifically includes: Convert the business consultation answer text corresponding to the business consultation template with the highest similarity to the business consultation text into an answer voice, and send the answer voice to the call center system, so that the call center system sends the answer voice Play to the client. 6. The business consultation processing method based on intelligent voice interaction technology according to claim 5, wherein the business consultation answer text corresponding to the business consultation template with the highest similarity with the business consultation text is converted into an answer voice, including: Use natural language processing method to extract text language features on the business consultation answer text; The extracted text language features are clustered using decision trees, and spectral coefficients and audio flags are obtained according to the clustering results and the hidden Markov model; According to the obtained spectral coefficients and audio flag bits, the maximum likelihood parameter generation algorithm is used to sample the spectral coefficients; The spectral coefficient sampling result is input into the vocoder to generate the speech waveform, and the answer speech is obtained. 7. The business consultation processing method based on intelligent voice interaction technology according to claim 1, is characterized in that, also comprises: If the highest similarity is less than the threshold, send a no-match answer instruction to the call center system, so that the call center system connects the customer's call to a manual customer service terminal according to the no-match answer instruction. 8. A business consultation processing device based on intelligent voice interaction technology, characterized in that, comprising: The business consultation voice acquisition unit is used to obtain the customer's business consultation voice; a voice recognition unit, configured to identify the business consultation text corresponding to the business consultation voice; a similarity matching unit, configured to perform similarity matching between the business consultation text and a preset business consultation template, and determine the business consultation template with the highest similarity with the business consultation text; A business consultation response information sending unit, configured to send the business consultation template with the highest similarity to the business consultation text if the similarity corresponding to the business consultation template with the highest similarity is greater than or equal to a preset threshold The business consultation response information corresponding to the consultation template is sent to the customer. 9. A computer device comprising a memory, a processor and a computer program stored on the memory and running on the processor, wherein the processor implements any one of claims 1 to 7 when the processor executes the computer program the steps of the method described in item. 10. A computer-readable storage medium on which a computer program/instruction is stored, characterized in that, when the computer program/instruction is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented. 11. A computer program product, comprising a computer program/instruction, characterized in that, when the computer program/instruction is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images