Typical skill recommendation method applied to seat questions and answers and related equipment

Abstract

The embodiment of the invention belongs to the technical field of artificial intelligence, and relates to a verbal skill recommendation method applied to seat questioning and answering and related equipment, and the method comprises the steps: receiving customer voice data collected by an audio collection terminal; performing voice translation operation on the client voice data to obtain a client voice text; performing semantic analysis operation on the client voice text to obtain client semantic information; reading a verbal skill database, and obtaining recommended verbal skill content corresponding to the customer semantic information from the verbal skill database; and outputting the recommended verbal skill content. According to the method and the device, the seat can be effectively helped to analyze the potential focus of attention of the customer, the scene that the seat is carried by himself or herself is avoided, the communication becomes more natural, efficient and smooth, the seat is helped to improve the communication skill, and meanwhile, the customer experience can be effectively improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent decision-making technology, applicable to the financial or medical fields, and in particular to a method for recommending speech techniques for agent question and answer and related equipment. Background Art

[0002] The emergence and popularization of the Internet has brought a large amount of information to users, satisfying their needs for information in the information age. However, it has also led to the fact that users are unable to obtain the information that is truly useful to them when faced with a large amount of information.

[0003] A very effective way to solve the above problems is the recommendation system, which is a personalized information recommendation system that recommends information and products of interest to users based on their information needs and interests. Most of the current information recommendation systems are based on generative large language models, which can generate human-like responses.

[0004] However, the applicant has found that traditional recommendation methods are unable to quickly identify customer intentions during real-time communication with customers, resulting in low accuracy of recommended content. Summary of the invention

[0005] The purpose of the embodiments of the present application is to propose a method and related equipment for recommending speech techniques for agent question and answer, so as to solve the problem of low accuracy of traditional recommendation methods.

[0006] In order to solve the above technical problems, the present application embodiment provides a method for recommending speech techniques for agent question and answer, which adopts the following technical solution:

[0007] Receive customer voice data collected by the audio collection terminal;

[0008] Performing a voice translation operation on the customer voice data to obtain a customer voice text;

[0009] Performing semantic analysis on the customer's voice text to obtain customer semantic information;

[0010] Reading a speech database, and obtaining recommended speech content corresponding to the customer semantic information in the speech database;

[0011] Output the recommended speech content.

[0012] Furthermore, the step of receiving the customer voice data collected by the audio collection terminal specifically includes the following steps:

[0013] The customer voice data is subjected to denoising processing to obtain clear customer voice data.

[0014] Furthermore, the step of performing semantic analysis on the customer voice text to obtain customer semantic information specifically includes the following steps:

[0015] Performing a deep semantic analysis operation on the customer's voice text according to a deep learning model to obtain key text information;

[0016] Perform semantic reasoning on the key text information according to a preset semantic rule library to obtain a knowledge graph;

[0017] Perform association analysis and reasoning on entities and relationships in the knowledge graph according to the graph neural network to obtain customer intentions;

[0018] The customer intention is classified according to a clustering algorithm to obtain the customer semantic information.

[0019] Furthermore, the step of reading the speech database and obtaining the recommended speech content corresponding to the customer semantic information in the speech database specifically includes the following steps:

[0020] Retrieving candidate speech related to the customer semantic information from the speech database;

[0021] According to information retrieval technology, the candidate speech words are sorted and screened for relevance to obtain a speech word subset;

[0022] The subset of speech words is screened to obtain the recommended speech words content.

[0023] Furthermore, the step of screening the speech subset to obtain the recommended speech content specifically includes the following steps:

[0024] According to the machine learning algorithm, the applicable scenarios and emotional tendencies of the speech subsets are respectively obtained;

[0025] The applicable scenarios and emotional tendencies of each subset of speech words are compared according to a collaborative filtering algorithm and customer personal information to obtain the recommended speech words content.

[0026] Furthermore, after the step of outputting the recommended speech content, the following steps are also included:

[0027] Obtaining feedback information corresponding to the content of the recommended speech sent by the customer service terminal;

[0028] The speech in the speech database is improved according to the feedback information.

[0029] In order to solve the above technical problems, the embodiment of the present application further provides a speech recommendation device for agent question and answer, which adopts the following technical solution:

[0030] A voice data acquisition module is used to receive the customer voice data collected by the audio collection terminal;

[0031] A voice translation module, used to perform voice translation operation on the customer voice data to obtain the customer voice text;

[0032] A semantic analysis module, used to perform semantic analysis on the customer's voice text to obtain customer semantic information;

[0033] A recommended speech acquisition module is used to read a speech database and acquire recommended speech content corresponding to the customer semantic information from the speech database;

[0034] The recommended speech output module is used to output the recommended speech content.

[0035] Furthermore, the semantic analysis module includes:

[0036] A deep semantic analysis submodule is used to perform a deep semantic analysis operation on the customer's voice text according to a deep learning model to obtain key text information;

[0037] A semantic reasoning submodule, used to perform semantic reasoning on the key text information according to a preset semantic rule library to obtain a knowledge graph;

[0038] A semantic reasoning submodule, which is used to perform association analysis and reasoning on entities and relationships in the knowledge graph based on a graph neural network to obtain customer intentions;

[0039] The classification submodule is used to classify the customer intention according to a clustering algorithm to obtain the customer semantic information.

[0040] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0041] It includes a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the method for recommending speech techniques applied to agent question and answer are implemented as described above.

[0042] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0043] The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the method for recommending speech techniques for agent question and answer as described above.

[0044] The present application provides a method for recommending speech techniques for agent question and answer, including: receiving customer voice data collected by an audio collection terminal; performing a voice translation operation on the customer voice data to obtain a customer voice text; performing a semantic analysis operation on the customer voice text to obtain customer semantic information; reading a speech technique database, and obtaining recommended speech techniques corresponding to the customer semantic information in the speech technique database; and outputting the recommended speech techniques. Compared with the prior art, the present application can effectively help agents analyze potential customer concerns, avoid scenarios where agents mechanically apply the techniques, make communication more natural, efficient, and smooth, help agents improve their communication skills, and effectively improve customer experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0046] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0047] Figure 2 It is a flowchart of the implementation of the method for recommending speech techniques for agent question and answer provided in an embodiment of the present application;

[0048] Figure 3 : is a structural diagram of a speech recommendation device for agent question and answer provided in an embodiment of the present application;

[0049] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0051] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0052] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0053] like Figure 1 As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0054] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0055] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV), a laptop computer, a desktop computer, etc.

[0056] The server 103 may be a server that provides various services, such as a background server that provides support for a web page displayed on the terminal device 101 .

[0057] It should be noted that the speech recommendation method applied to the agent question and answer provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the speech recommendation device applied to the agent question and answer is generally set in the server / terminal device.

[0058] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0059] Continue to refer Figure 2 , showing a flow chart of an embodiment of a method for recommending speech techniques for agent question and answer according to the present application. The method for recommending speech techniques for agent question and answer comprises: step S201, step S202, step S203, step S204 and step S205.

[0060] In step S201, receiving customer voice data collected by an audio collection terminal;

[0061] In step S202, a voice translation operation is performed on the customer voice data to obtain the customer voice text;

[0062] In step S203, semantic analysis is performed on the customer's voice text to obtain customer semantic information;

[0063] In step S204, the speech database is read, and the recommended speech content corresponding to the customer semantic information is obtained in the speech database;

[0064] In step S205, the recommended speech content is output.

[0065] In an embodiment of the present application, the customer voice data may be "transaction voice data or payment voice data or business voice data or purchase voice data" related to a financial institution (such as a bank, etc.). The customer voice data may also be medical voice data related to medical scenarios, such as personal health records, prescriptions, examination reports and other voice data. It should be understood that the examples of customer voice data here are only for the convenience of understanding and are not used to limit the present application.

[0066] In the embodiment of the present application, the key information of the communication between the agent and the user is compared with the speech pool in real time, and the optimal speech solution is screened out and fed back to the agent to realize the speech communication closed loop. The specific implementation steps are as follows:

[0067] ① Select the call recordings of agents with excellent order placement rates from the massive call recordings of tens of thousands of agents every day, and upload these recordings to the big data analysis platform for data analysis;

[0068] ② The big data analysis platform converts the recording files through file streams, converts the agent recording files into text mode, and then cleans the massive text files, identifying the top 10 excellent speech cases through the set big data analysis model;

[0069] ③Distribute excellent speech cases to agents in the form of courses for learning. Through centralized or fragmented learning, agents’ speech skills for placing orders can be strengthened and their daily communication skills can be improved;

[0070] ④ The above are all based on the existing speech schemes. The following is the most important optimization of the existing model technology in this scheme. Through this scheme, the three-party linkage of seats, customers, and machines (speech pool) is realized. The specific optimization plan is: during the real-time communication with customers, the seats can quickly extract and grasp the key points of customer concern through machine voice translation. Then, the extracted keyword information is quickly matched with the relevant successful case speech information in the speech pool. For example, if the user is concerned about the price, then the seat is provided with price-related optimized speech. If the customer is concerned about after-sales service, then after-sales related services are provided in real time. If the customer is concerned about the product itself, then the seat is asked to focus on introducing the advantages of the product. In this way, a communication closed loop is formed among the seat, the customer and the speech pool. Through this communication mode, the above-mentioned rigid scenarios are avoided, making communication more natural, efficient and smooth.

[0071] In an embodiment of the present application, a method for recommending speech techniques for agent question and answer is provided, including: receiving customer voice data collected by an audio collection terminal; performing a voice translation operation on the customer voice data to obtain a customer voice text; performing a semantic analysis operation on the customer voice text to obtain customer semantic information; reading a speech technique database, and obtaining recommended speech techniques corresponding to the customer semantic information in the speech technique database; and outputting the recommended speech techniques. Compared with the prior art, the present application can effectively help agents analyze potential customer concerns, avoid scenarios where agents mechanically apply the techniques, make communication more natural, efficient, and smooth, help agents improve their communication skills, and effectively improve customer experience.

[0072] In some optional implementations of the embodiments of the present application, the step of receiving the customer voice data collected by the audio collection terminal specifically includes the following steps:

[0073] The customer voice data is denoised to obtain clear customer voice data.

[0074] In the embodiments of the present application, denoising can be performed using a variety of methods and techniques, including but not limited to:

[0075] (1) Filtering technology: Remove noise within a specific frequency range by designing filters.

[0076] (2) Signal enhancement algorithms: such as adaptive filtering and Kalman filtering. These algorithms can be dynamically adjusted according to the statistical characteristics of speech signals and noise to achieve the best denoising effect.

[0077] (3) Machine learning technology: In recent years, machine learning technologies such as deep learning have also been widely used in the field of speech denoising. By training models to identify and remove noise, this method performs well in dealing with complex noise environments.

[0078] (4) Speech enhancement software: There are many professional speech enhancement software and services on the market. They provide user-friendly interfaces and powerful denoising functions, making it easy for non-professionals to perform speech denoising.

[0079] In the embodiment of the present application, after the above denoising process, clearer customer voice data can be obtained. Such data not only improves the audibility and comprehensibility of the voice, but also provides more accurate and reliable input for subsequent voice recognition, voice analysis and other technologies.

[0080] In some optional implementations of the embodiments of the present application, the above-mentioned step of performing a semantic analysis operation on the customer's voice text to obtain the customer's semantic information specifically includes the following steps:

[0081] Perform deep semantic analysis on customer voice texts based on deep learning models to obtain key text information;

[0082] Perform semantic reasoning on key text information based on the preset semantic rule base to obtain a knowledge graph;

[0083] Perform association analysis and reasoning on entities and relationships in the knowledge graph based on graph neural networks to obtain customer intent;

[0084] Customer intentions are classified according to the clustering algorithm to obtain customer semantic information.

[0085] In the embodiments of the present application, deep learning models (such as BERT, GPT, etc.) can identify keywords and phrases in the text and the semantic relationships between them, thereby extracting key text information that is crucial for subsequent analysis. This information may include specific products, services, questions or needs mentioned by customers.

[0086] In an embodiment of the present application, after extracting the key text information, the system uses a preset semantic rule base for further semantic reasoning. These rule bases contain knowledge in specific fields, common sense reasoning rules, logical relationships, etc. By applying these rules, the system can organize the key text information into a structured knowledge graph. A knowledge graph is a graphical data structure that can clearly represent entities (such as customers, products, etc.) and the relationships between them (such as purchases, consultations, etc.).

[0087] In an embodiment of the present application, after obtaining the knowledge graph, the entities and relationships in the graph are subjected to association analysis and reasoning based on a graph neural network (GNN). GNN can capture the complex interactions between nodes (entities) and edges (relationships) in the graph, and update the representation of the nodes by iteratively propagating information. It can reflect the position of the node in the graph, the information of the neighboring nodes, and the relationship between them. Through the reasoning of GNN, the system can more deeply understand the customer's intentions, that is, the goals that the customer wants to achieve or the problems that they want to solve.

[0088] In an embodiment of the present application, the system uses a clustering algorithm to classify customer intent. The clustering algorithm can classify the intents into different categories or groups based on the similarities between them. These categories represent different customer needs, problem types, or business scenarios. Through classification, the system can organize the customer's semantic information in a more orderly and understandable way, providing support for subsequent business decisions, customer service, or personalized recommendations.

[0089] In some optional implementations of the embodiments of the present application, the above-mentioned step of reading the speech database and obtaining the recommended speech content corresponding to the customer semantic information in the speech database specifically includes the following steps:

[0090] Retrieve candidate dialogues related to customer semantic information from the dialogue database;

[0091] According to information retrieval technology, candidate speech words are sorted and screened for relevance to obtain a subset of speech words;

[0092] Filter the subset of dialogues to obtain recommended dialogue content.

[0093] In the embodiment of the present application, a pre-built speech database is first accessed. This database contains a large number of carefully designed and optimized dialogue speeches, designed to cope with various possible customer scenarios and needs. These speeches cover product introduction, service consultation, problem solving, complaint handling and other aspects; in order to find candidate speeches related to customer semantic information, the system uses information retrieval (IR) technology. Specifically, it will use customer semantic information (such as customer intent, key text information, etc.) as query conditions and match each speech in the speech database. The matching process may consider multiple factors such as keywords, phrases, semantic similarity, etc. in the speech. Through this step, the system can preliminarily screen out a set of candidate speeches related to customer semantic information, providing a basis for subsequent relevance sorting and screening operations.

[0094] In an embodiment of the present application, after obtaining the candidate words, the system will use more sophisticated information retrieval technology to sort and screen them for relevance. This step is intended to further narrow the scope of candidate words and ensure that the words finally selected are highly relevant and effective to the customer's semantic information. Relevance sorting will be based on a variety of factors, such as the semantic similarity between the words and the customer's semantic information, the number of keywords contained in the words, the historical use effect of the words in the database, etc. The system will use machine learning models (such as text matching models) to evaluate these factors and give a relevance score for each candidate word. After obtaining the relevance score, the system will screen the candidate words according to preset thresholds or sorting rules. Specifically, it will select the top N words with the highest scores, or select words with scores exceeding a certain threshold as a subset of words.

[0095] In an embodiment of the present application, after obtaining the subset of words, the system needs to perform further screening operations to ensure that the final recommended words content meets both customer needs and the business specifications and service standards of the enterprise. This step will consider multiple factors, such as the applicability of the words (whether it is applicable to the current customer scenario), the compliance of the words (whether it complies with laws, regulations and corporate policies), the degree of personalization of the words (whether the customer's personalized needs and preferences are taken into account), etc. The system will use a rule engine or machine learning model to evaluate these factors and further screen the subset of words. Ultimately, through this series of screening operations, the system can select one or more words that best suit the current customer scenario and needs as recommended words. These recommended words will be used directly for dialogue and communication with customers to provide high-quality services and solutions.

[0096] In some optional implementations of the embodiments of the present application, the step of screening the above-mentioned dialogue subset to obtain the recommended dialogue content specifically includes the following steps:

[0097] According to the machine learning algorithm, the applicable scenarios and emotional tendencies of the sub-sets of the speech are obtained respectively;

[0098] Based on the collaborative filtering algorithm and customer personal information, the applicable scenarios and emotional tendencies of each script subset are compared to obtain recommended script content.

[0099] In an embodiment of the present application, in this step, the system uses a machine learning algorithm to conduct an in-depth analysis of a subset of dialogues to obtain the applicable scenarios and emotional tendencies of each dialogue. For applicable scenarios, the system uses classification algorithms (such as decision trees, random forests, support vector machines, etc.) to label and classify dialogues. These algorithms divide the dialogues into different applicable scenarios, such as product consultation, service complaints, after-sales support, etc., based on information such as keywords, phrases, and semantic features in the dialogues. Through this step, the system can understand which customer scenarios each dialogue is most suitable for.

[0100] In the embodiment of the present application, for emotional tendency, the system uses sentiment analysis algorithms (such as sentiment classification models based on deep learning) to evaluate the emotional color expressed by the speech. These algorithms can identify information such as emotional vocabulary, tone and intonation in the speech, so as to determine whether the speech is positive, negative or neutral. Through the analysis of emotional tendency, the system can better understand the emotional impact that the speech may have on customers.

[0101] In an embodiment of the present application, after obtaining the applicable scenarios and emotional tendencies of a subset of words, the system uses a collaborative filtering algorithm and customer personal information to compare and screen these words. The collaborative filtering algorithm is a recommendation algorithm based on user behavior or preference, which can predict the content that the user may be interested in by analyzing the user's historical data. In this scenario, the system uses customer personal information (such as age, gender, purchase history, service needs, etc.) as input, and combines the collaborative filtering algorithm to evaluate the degree of match between each word and customer needs. At the same time, the system will also consider the applicable scenarios and emotional tendencies of the words to ensure that the recommended words not only meet the customer's personal information and needs, but also communicate with the customer in an appropriate emotional expression in the appropriate scenario.

[0102] In the embodiment of the present application, through the collaborative filtering algorithm and the comparison and analysis of the customer's personal information, the system can screen out the speech content that best meets the customer's needs and expectations. These recommended speech will be comprehensively considered and selected based on multiple factors such as the customer's personal information, service needs, applicable scenarios, and emotional tendencies. Finally, the system will present the recommended speech content to the customer service staff or the intelligent customer service system for use in conversations with customers. These recommended speech can not only improve the efficiency and accuracy of the conversation, but also enhance customer satisfaction and loyalty.

[0103] In some optional implementations of the embodiments of the present application, after the above step of outputting the recommended speech content, the following steps are also included:

[0104] Obtaining feedback information corresponding to the recommended speech content sent by the customer service terminal;

[0105] Improve the dialogue in the dialogue database based on the feedback information.

[0106] In the embodiment of the present application, the system will receive feedback information from the customer service terminal (such as the customer service staff's computer, the interface of the intelligent customer service system, etc.). This feedback information corresponds to the content of the speech recommended by the system before, and is intended to reflect the effect of the speech in actual application and customer feedback.

[0107] In the embodiment of the present application, after obtaining the feedback information, the system will analyze and process the information to identify the problems and improvement points in the speech. Then, based on the analysis results, the system will perform improvement operations on the speech in the speech database, where the improvement operation includes the following aspects:

[0108] ①Correct errors: Correct grammatical errors, unclear semantics, inaccurate information and other issues in the speech.

[0109] ② Optimize expression: Based on customer feedback and customer service evaluation, optimize the expression of dialogue to make it more concise, clear and easy to understand.

[0110] ③ Increase scenario coverage: Based on the customer scenarios and needs mentioned in the feedback information, add new words or dialogues to expand and cover more customer scenarios.

[0111] ④ Adjust emotional tendencies: Based on the customer’s emotional feedback, adjust the emotional tendencies of the dialogue techniques to make them more in line with the customer’s expectations and feelings.

[0112] In the embodiment of the present application, the dialogue database can be continuously iterated and optimized to improve the quality and applicability of the dialogue.

[0113] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0114] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0115] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0116] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0117] Further references Figure 3 , as a response to the above Figure 2 In order to realize the method shown in the figure, the present application provides an embodiment of a speech recommendation device for agent question and answer, and the device embodiment is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0118] like Figure 3 As shown, the speech technique recommendation device 200 applied to the agent question and answer of the embodiment of the present application includes:

[0119] The voice data acquisition module 210 is used to receive the customer voice data collected by the audio collection terminal;

[0120] The voice translation module 220 is used to perform voice translation operation on the customer voice data to obtain the customer voice text;

[0121] Semantic analysis module 230, used to perform semantic analysis on customer voice text to obtain customer semantic information;

[0122] The recommended speech acquisition module 240 is used to read the speech database and acquire the recommended speech content corresponding to the customer semantic information in the speech database;

[0123] The recommended speech output module 250 is used to output the recommended speech content.

[0124] In the embodiment of the present application, a speech recommendation device 200 for agent question and answer is provided, including: a voice data acquisition module 210, for receiving customer voice data collected by an audio collection terminal; a voice translation module 220, for performing a voice translation operation on the customer voice data to obtain a customer voice text; a semantic analysis module 230, for performing a semantic analysis operation on the customer voice text to obtain customer semantic information; a recommended speech acquisition module 240, for reading a speech database, and obtaining recommended speech content corresponding to the customer semantic information in the speech database; a recommended speech output module 250, for outputting the recommended speech content. Compared with the prior art, the present application can effectively help agents analyze the potential focus of customers, avoid the scene of agents copying and pasting, make communication more natural, efficient and smooth, help agents improve their communication skills, and at the same time, can effectively improve customer experience.

[0125] In some optional implementations of the embodiments of the present application, the semantic analysis module includes:

[0126] The deep semantic analysis submodule is used to perform deep semantic analysis on the customer's voice text according to the deep learning model to obtain key text information;

[0127] The semantic reasoning submodule is used to perform semantic reasoning on key text information according to the preset semantic rule base to obtain a knowledge graph;

[0128] The semantic reasoning submodule is used to perform association analysis and reasoning on entities and relationships in the knowledge graph based on the graph neural network to obtain customer intent;

[0129] The classification submodule is used to classify customer intentions according to the clustering algorithm to obtain customer semantic information.

[0130] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of a computer device according to an embodiment of the present application.

[0131] The computer device 300 includes a memory 310, a processor 320, and a network interface 330 that are interconnected and communicated through a system bus. It should be noted that the figure only shows a computer device 300 having components 310-330, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital signal processors (Digital Signal Processor, DSP), embedded devices, etc.

[0132] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0133] The memory 310 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 310 can be an internal storage unit of the computer device 300, such as a hard disk or memory of the computer device 300. In other embodiments, the memory 310 can also be an external storage device of the computer device 300, such as a plug-in hard disk equipped on the computer device 300, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card, etc. Of course, the memory 310 can also include both the internal storage unit of the computer device 300 and its external storage device. In the embodiment of the present application, the memory 310 is generally used to store the operating system and various application software installed on the computer device 300, such as computer-readable instructions for the speech recommendation method for agent question and answer, etc. In addition, the memory 310 can also be used to temporarily store various data that have been output or are to be output.

[0134] The processor 320 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments. The processor 320 is generally used to control the overall operation of the computer device 300. In the embodiment of the present application, the processor 320 is used to run the computer-readable instructions stored in the memory 310 or process data, such as running the computer-readable instructions of the method for recommending speech techniques for agent question and answer.

[0135] The network interface 330 may include a wireless network interface or a wired network interface. The network interface 330 is generally used to establish a communication connection between the computer device 300 and other electronic devices.

[0136] The computer equipment provided in this application can effectively help agents analyze potential customer concerns, avoid scenarios where agents follow a mechanical approach, make communication more natural, efficient, and smooth, help agents improve their communication skills, and at the same time, effectively improve customer experience.

[0137] The present application also provides another implementation, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the method for recommending speech techniques applied to agent question and answer as described above.

[0138] The computer-readable storage medium provided in this application can effectively help agents analyze potential customer concerns, avoid scenarios where agents follow a mechanical approach, make communication more natural, efficient, and smooth, help agents improve their communication skills, and at the same time, effectively improve customer experience.

[0139] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0140] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A method for recommending speech techniques for agent question and answer, characterized in that: The steps include: Receive customer voice data collected by the audio collection terminal; Performing a voice translation operation on the customer voice data to obtain a customer voice text; Performing semantic analysis on the customer's voice text to obtain customer semantic information; Reading a speech database, and obtaining recommended speech content corresponding to the customer semantic information in the speech database; Output the recommended speech content.

2. The method for recommending speech techniques for agent question and answer according to claim 1, characterized in that: The step of receiving the customer voice data collected by the audio collection terminal specifically includes the following steps: The customer voice data is subjected to denoising processing to obtain clear customer voice data.

3. The method for recommending speech techniques for agent question and answer according to claim 1, characterized in that: The step of performing semantic analysis on the customer voice text to obtain customer semantic information specifically includes the following steps: Performing a deep semantic analysis operation on the customer's voice text according to a deep learning model to obtain key text information; Perform semantic reasoning on the key text information according to a preset semantic rule library to obtain a knowledge graph; Perform association analysis and reasoning on entities and relationships in the knowledge graph according to the graph neural network to obtain customer intentions; The customer intention is classified according to a clustering algorithm to obtain the customer semantic information.

4. The method for recommending speech techniques for agent question and answer according to claim 1, characterized in that: The step of reading the speech database and obtaining the recommended speech content corresponding to the customer semantic information in the speech database specifically includes the following steps: Retrieving candidate speech related to the customer semantic information from the speech database; According to information retrieval technology, the candidate speech words are sorted and screened for relevance to obtain a speech word subset; The subset of speech words is screened to obtain the recommended speech words content.

5. The method for recommending speech techniques for agent question and answer according to claim 4, characterized in that: The step of screening the speech subset to obtain the recommended speech content specifically includes the following steps: According to the machine learning algorithm, the applicable scenarios and emotional tendencies of the speech subsets are respectively obtained; The applicable scenarios and emotional tendencies of each subset of speech words are compared according to a collaborative filtering algorithm and customer personal information to obtain the recommended speech words content.

6. The method for recommending speech techniques for agent question and answer according to claim 1, characterized in that: After the step of outputting the recommended speech content, the following steps are also included: Obtaining feedback information corresponding to the content of the recommended speech sent by the customer service terminal; The speech in the speech database is improved according to the feedback information.

7. A speech recommendation device for agent question and answer, characterized in that: include: A voice data acquisition module is used to receive the customer voice data collected by the audio collection terminal; A voice translation module, used to perform voice translation operation on the customer voice data to obtain the customer voice text; A semantic analysis module, used to perform semantic analysis on the customer's voice text to obtain customer semantic information; A recommended speech acquisition module is used to read a speech database and acquire recommended speech content corresponding to the customer semantic information from the speech database; The recommended speech output module is used to output the recommended speech content.

8. The speech recommendation device for agent question and answer according to claim 7, characterized in that: The semantic analysis module includes: A deep semantic analysis submodule is used to perform a deep semantic analysis operation on the customer's voice text according to a deep learning model to obtain key text information; A semantic reasoning submodule, used to perform semantic reasoning on the key text information according to a preset semantic rule library to obtain a knowledge graph; A semantic reasoning submodule, which is used to perform association analysis and reasoning on entities and relationships in the knowledge graph based on a graph neural network to obtain customer intentions; The classification submodule is used to classify the customer intention according to a clustering algorithm to obtain the customer semantic information.

9. A computer device comprising a memory and a processor, characterized in that: The memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the method for recommending speech techniques for agent question and answer as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the method for recommending speech techniques for agent question and answer as described in any one of claims 1 to 6.

Images