Service dialogue quality inspection model training method and device, equipment and storage medium

Abstract

The invention provides a service dialogue quality inspection model training method and device, equipment and a storage medium. The invention relates to the technical field of artificial intelligence, in particular to the fields of large models, instruction optimization, data analysis and the like, and can be applied to application scenes such as intelligent customer service, satisfaction management, dialogue analysis, dialogue quality inspection and the like. According to the specific implementation scheme, a target prompt text is obtained; inputting the service dialogue content and the target prompt text into a teacher model, so that the teacher model generates a preliminary analysis result of the service dialogue content based on the target prompt text; and inputting the preliminary analysis result into a student model, so that the student model performs learning and optimization according to the preliminary analysis result: determining the student model after learning and optimization as a service dialogue quality inspection model, and the service dialogue quality inspection model is used for generating a service dialogue quality inspection result according to the target prompt text.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to the fields of large models, instruction optimization, data analysis, etc., and can be used in application scenarios such as intelligent customer service, satisfaction management, dialogue analysis, dialogue quality inspection, etc., and specifically relates to a training method, device, equipment and storage medium for a service dialogue quality inspection model. Background Art

[0002] Existing customer conversation quality inspection technologies mainly rely on manual annotation rules or statistical-based machine learning models. However, with the increasing complexity of conversation content and the diversification of customer needs, traditional quality inspection systems are difficult to adapt and respond quickly, especially when dealing with unknown conversations or unexpected problems. Summary of the invention

[0003] The present invention provides a training method, device, equipment and storage medium for a service dialogue quality inspection model.

[0004] According to a first aspect of the present disclosure, a training method for a service conversation quality inspection model is provided, the method comprising: obtaining a target prompt text; inputting the service conversation content and the target prompt text into a teacher model, so that the teacher model generates a preliminary analysis result of the service conversation content based on the target prompt text; inputting the preliminary analysis result into a student model, so that the student model performs learning and optimization according to the preliminary analysis result; determining the student model after learning and optimization as a service conversation quality inspection model, and the service conversation quality inspection model is used to generate a service conversation quality inspection result according to the target prompt text.

[0005] According to a second aspect of the present disclosure, a large model-based service conversation quality inspection method is provided, the method comprising: obtaining the service conversation content to be quality inspected, inputting the service conversation content to be quality inspected into a service conversation quality inspection model, and obtaining the service conversation quality inspection result output by the service conversation quality inspection model based on the target prompt text; wherein the service conversation quality inspection model is obtained by training the method of the first aspect mentioned above.

[0006] According to a third aspect of the present disclosure, a training device for a service dialogue quality inspection model is provided, the device comprising: a first acquisition module, for acquiring a target prompt text; a first control module, for inputting the service dialogue content and the target prompt text into a teacher model, so that the teacher model generates a preliminary analysis result of the service dialogue content based on the target prompt text; a second control module, for inputting the preliminary analysis result into a student model, so that the student model performs learning and optimization according to the preliminary analysis result; a determination module, for determining the student model after learning and optimization as a service dialogue quality inspection model, the service dialogue quality inspection model being used to generate a service dialogue quality inspection result according to the target prompt text.

[0007] According to a fourth aspect of the present disclosure, a service conversation quality inspection device based on a large model is provided, the device comprising: a second acquisition module, used to obtain the service conversation content to be quality inspected; a quality inspection module, used to input the service conversation content to be quality inspected into a service conversation quality inspection model, and obtain a service conversation quality inspection result output by the service conversation quality inspection model based on a target prompt text; wherein the service conversation quality inspection model is obtained by training using the method of the first aspect.

[0008] According to a fifth aspect of the present disclosure, there is provided an electronic device, including:

[0009] at least one processor; and

[0010] a memory communicatively connected to the at least one processor; wherein,

[0011] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute any method in the embodiments of the present disclosure.

[0012] According to a sixth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute any method according to the embodiments of the present disclosure.

[0013] According to a seventh aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements any method according to the embodiments of the present disclosure when executed by a processor.

[0014] The solution disclosed in the present invention can enhance the generalization ability of the model and improve the accuracy and efficiency of dialogue quality inspection.

[0015] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0017] Figure 1 : is a flow chart of a training method for a service dialogue quality inspection model according to an embodiment of the present disclosure.

[0018] Figure 2 Schematic diagram of target prompt text according to an embodiment of the present disclosure.

[0019] Figure 3 Schematic diagram of the evolution process of the service dialogue quality inspection model according to the embodiment of the present disclosure.

[0020] Figure 4 Schematic diagram of an example of a service dialogue quality inspection model according to an embodiment of the present disclosure.

[0021] Figure 5 is a flowchart of a service dialogue quality inspection method based on a large model according to an embodiment of the present disclosure;

[0022] Figure 6 : is a structural diagram of a training device for a service dialogue quality inspection model according to an embodiment of the present disclosure.

[0023] Figure 7 Schematic diagram of the structure of the service dialogue quality inspection device based on the big model according to the embodiment of the present disclosure.

[0024] Figure 8 Schematic diagram of a scenario of a training method for a service dialogue quality inspection model according to an embodiment of the present disclosure.

[0025] Fig. 9 Schematic diagram of a scenario of a service dialogue quality inspection method based on a large model according to an embodiment of the present disclosure.

[0026] Fig.10 It is a structural schematic diagram of an electronic device used to implement the training method of the service dialogue quality inspection model and / or the service dialogue quality inspection method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present disclosure. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0028] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. The term "at least one" in this article means any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C. The terms "first" and "second" in this article refer to multiple similar technical terms and distinguish them. They do not mean to limit the order or to limit them to only two. For example, the first feature and the second feature refer to two types / two features. The first feature can be one or more, and the second feature can also be one or more.

[0029] In addition, in order to better illustrate the present disclosure, numerous specific details are given in the following specific embodiments. It should be understood by those skilled in the art that the present disclosure can also be implemented without certain specific details. In some examples, methods, means, components and circuits well known to those skilled in the art are not described in detail in order to highlight the subject matter of the present disclosure.

[0030] Before introducing the technical solutions of the embodiments of the present disclosure, the technical terms that may be used in the present disclosure are further explained:

[0031] Teacher model: A pre-trained model with strong analytical capabilities that is used to generate preliminary analysis results of service conversation content.

[0032] Student model: The quality inspection model to be trained, which learns and optimizes by imitating the output of the teacher model.

[0033] Target prompt text: text information related to quality inspection standards, business specifications, or user expectations, used to guide the teacher model and student model to analyze the conversation content.

[0034] In related technologies, conversation quality inspection systems usually use predefined rule sets or models to analyze the content of the conversation and make judgments. Many quality inspection systems can only rely on fixed model structures and cannot improve the quality inspection effect through the evolution of the model itself. Conversation quality inspection technology has the following main disadvantages: it relies on predefined rules and pre-trained models, and cannot automatically adapt to new conversation modes or changes in customer needs, resulting in insufficient accuracy of conversation quality inspection. Most quality inspection systems cannot improve the performance of conversation quality inspection through self-learning or evolutionary models, and need to frequently rely on manual intervention to update the model, which is inefficient. When it involves multiple rounds of conversations or complex problems, the system's classification and judgment capabilities are weak, and it is easy to make misjudgments or omissions.

[0035] In order to at least partially solve one or more of the above-mentioned problems and other potential problems, the present disclosure proposes a model-based service dialogue quality inspection method, which can improve the accuracy and efficiency of dialogue quality inspection through self-learning and model evolution, solves the problems of low efficiency, inaccurate judgment and lack of evolution capability existing in the dialogue quality inspection system in the current customer service field, and can adapt to changing customer needs and complex dialogue scenarios.

[0036] The present disclosure provides a method for training a service dialogue quality inspection model. Figure 1It is a flowchart of a training method for a service dialogue quality inspection model according to an embodiment of the present disclosure. The training method for a service dialogue quality inspection model can be applied to a service dialogue quality inspection model training device. The service dialogue quality inspection model training device is located in an electronic device. The electronic device includes but is not limited to fixed devices and / or mobile devices. For example, fixed devices include but are not limited to servers, and servers can be cloud servers or ordinary servers. For example, mobile devices include but are not limited to: mobile phones, tablet computers. In some possible implementations, the training method for the service dialogue quality inspection model can also be implemented by a processor calling computer-readable instructions stored in a memory. For example Figure 1 As shown, the training method of the service dialogue quality inspection model includes:

[0037] S101. Get target prompt text

[0038] S102, inputting the service dialogue content and the target prompt text into the teacher model, so that the teacher model generates a preliminary analysis result of the service dialogue content based on the target prompt text;

[0039] S103, inputting the preliminary analysis results into the student model, so that the student model can learn and optimize according to the preliminary analysis results;

[0040] S104. Determine the student model after learning and optimization as a service dialogue quality inspection model, and the service dialogue quality inspection model is used to generate a service dialogue quality inspection result according to the target prompt text.

[0041] In the disclosed embodiment, the target prompt text includes text for guiding the model to perform analysis, which contains key information and requirements required for the analysis.

[0042] In the disclosed embodiment, the teacher model is a pre-trained model with analytical capabilities, which can analyze the service dialogue content according to the target prompt text.

[0043] In the disclosed embodiment, the student model is a pre-trained quality inspection model that can learn and optimize based on the preliminary analysis results of the teacher model.

[0044] In the disclosed embodiment, the target prompt text includes key information of multiple dimensions, and the multiple dimensions include: problem information, problem status, provided solutions, solution results and service evaluation.

[0045] In some embodiments, after the service conversation content and target prompt text are input into the teacher model, the teacher model will generate a preliminary analysis result, which may include key information extraction, sentiment tendency judgment, compliance check and other levels of content in the conversation.

[0046] In some embodiments, the student model learns and optimizes based on the preliminary analysis results, including: by imitating the output of the teacher model, the student model gradually masters the ability to analyze the content of the conversation, identify potential problems, and generate quality inspection results. At the same time, the student model can also continuously optimize its internal parameters and strategies through self-iteration and feedback mechanisms to improve the accuracy and efficiency of quality inspection.

[0047] In some embodiments, the student model after learning and optimization is determined as the service dialogue quality inspection model, including: after multiple rounds of learning and optimization, the performance of the student model will gradually stabilize and reach the expected quality inspection standards, and the student model can be determined as the final service dialogue quality inspection model and deployed in actual application scenarios.

[0048] For example, suppose the customer service department of an e-commerce platform needs to conduct quality inspection on the conversations between customer service and users. Collect and organize the historical conversation data between customer service and users, as well as the service specifications and user expectations within the enterprise, to form a target prompt text set; input the historical conversation data and target prompt text into the teacher model to generate preliminary analysis results; input the preliminary analysis results into the student model, and learn and optimize by imitating the output of the teacher model; after multiple rounds of learning and optimization, determine the student model as the final service conversation quality inspection model and deploy it to the customer service system of the e-commerce platform. When the customer service has a conversation with the user, the quality inspection model will automatically conduct quality inspection on the conversation content and generate quality inspection results. If there are potential problems or non-compliance with service specifications in the conversation, the quality inspection model will issue warnings or prompts in time so that the customer service staff can correct them in time.

[0049] The technical solution of the disclosed embodiment uses the teacher model for preliminary analysis, which can greatly reduce the computational burden of the student model and improve the real-time and accuracy of quality inspection. By imitating the output of the teacher model, the student model can learn more about conversation content analysis, thereby enhancing its ability to handle different scenarios and problems. The trained service conversation quality inspection model is deployed in actual applications, which can automatically complete quality inspection tasks, greatly reduce the workload and cost of manual review, and improve quality inspection efficiency.

[0050] In some embodiments, obtaining the target prompt text includes: obtaining an initial prompt text, optimizing and adjusting the initial prompt text according to the needs of the service industry, and obtaining the target prompt text.

[0051] In the disclosed embodiment, the target prompt text is used to guide the model to generate a specific output. It contains information and instructions that the model needs to understand and is an important basis for the model to generate output.

[0052] In the disclosed embodiments, the service industry refers to an industry or department that provides service labor, which has a wide range and includes catering, tourism, finance, medical care, education, e-commerce and other fields.

[0053] In the disclosed embodiment, the initial prompt text is a basic, unoptimized text, which may come from a preset template, historical data or other related resources. These texts usually contain some basic instructions or information, but have not been customized for the needs of a specific service industry.

[0054] In some embodiments, optimization and adjustment are performed according to the needs of the service industry, including: first, an in-depth analysis of the needs of the service industry is required. This includes understanding the specific scenarios, business processes, user preferences, and expected outputs of the service industry. Through demand analysis, it can be clarified what key information and instructions the target prompt text needs to contain. Based on the results of the demand analysis, the initial prompt text is optimized and adjusted. This may include adding necessary contextual information, adjusting the way the instructions are expressed, deleting redundant information, etc. The goal of the optimization is to make the target prompt text more in line with the actual needs of the service industry and to guide the model to generate more accurate and useful outputs.

[0055] In some embodiments, the initial prompt text and related background information are input into the big model. The optimization suggestions are generated by using the natural language processing capabilities and understanding of a large amount of text data of the big model. The initial prompt text is optimized and adjusted based on the optimization suggestions to obtain the target prompt text. Here, the adjustment may include adjusting the wording, sentence pattern, structure, etc. to make the prompt text clearer, more accurate, and more attractive. Here, the background information may include the background, characteristics, customer needs, etc. of the service industry, so that this information can be incorporated into the optimization process.

[0056] In some implementations, after the optimization is completed, the target prompt text needs to be tested and verified. For example, the target prompt text is input into the model to observe whether the output of the model meets expectations. If the output of the model does not meet expectations, the target prompt text may need to be further adjusted.

[0057] Suppose a hotel industry wants to use a big model to answer employees' questions about the hotel property management system (PMS). The initial prompt text may be a simple inquiry: "Please explain what PMS is?" However, such a prompt text may not provide the specific information needed by employees. In order to optimize this prompt text, we can make the following adjustments: Add contextual information: Adjust the inquiry to: "In the tourism industry, what does PMS usually refer to?" Such a statement can provide employees with more background information and help them better understand the concept of PMS. Further adjust the prompt text to: "Please explain what PMS is in the tourism industry in a concise and clear way, and give its main functions." Such a statement clarifies the output requirements of the model and helps to generate more accurate and useful answers. Through the above optimization and adjustment, a target prompt text that better meets the needs of the hotel industry can be obtained, thereby guiding the model to generate more accurate and useful outputs.

[0058] In this way, by optimizing and adjusting the target prompt text, it can be made more in line with the actual needs of the service industry, thereby improving the accuracy of the model output; the optimized target prompt text can more effectively guide the model to output, thereby reducing unnecessary iterations and adjustments and improving work efficiency.

[0059] In some embodiments, the target prompt text includes key information of multiple dimensions, and the multiple dimensions include: problem information, problem status, provided solutions, solution results, and service evaluation.

[0060] In the disclosed embodiment, the problem information is the core problem or demand extracted from the conversation, which clarifies the specific matters that the user wants to solve, and may include detailed information such as the description of the problem, the time and place of occurrence, etc.

[0061] In the disclosed embodiment, the problem status is to identify the current status of the problem, such as reported, being processed, resolved, etc. The problem status reflects the change of the problem during the dialogue process, such as from unknown to known, from pending to processed, etc.

[0062] In the embodiments of the present disclosure, the solution provided is a list of solutions or suggestions proposed for the problem, including the specific content of the solution, implementation steps, expected effects, etc.

[0063] In the disclosed embodiment, the solution result is to record the actual result after the implementation of the solution, such as whether the problem is solved and the degree of solution, etc., which reflects the effectiveness of the solution and provides a basis for subsequent improvements.

[0064] In the disclosed embodiment, the service evaluation is to collect user feedback on the service process and results, including evaluations in terms of satisfaction, service quality, response speed, etc.

[0065] When constructing the target prompt text, a templated approach can be used to provide a fixed format and expression for each dimension. At the same time, the template can be appropriately adjusted and supplemented according to the specific content of the conversation and the characteristics of the service industry to ensure the accuracy and completeness of the target prompt text.

[0066] Figure 2 A schematic diagram of the target prompt text is shown, such as Figure 2 As shown, the target prompt text includes task requirement description, extracted problem information, solution, solution status, solution result, service evaluation and other contents.

[0067] Specifically, the task requirements are described as: To understand the content of the conversation, you need to read and understand the content of the conversation between the customer and the human customer service.

[0068] Specifically, problem information is extracted, including problem description (description): identifying the key description of the problem or demand raised by the customer; phenomenon description (phenomenon): recording the specific phenomenon or situation encountered by the customer; classification label (classification): accurately selecting the label that best fits the problem description or phenomenon from the given classification list, and if necessary, supplemented by solution content judgment to ensure that the label belongs to the given list.

[0069] Specifically, the resolution includes: paying attention to all solutions provided by customer service and describing each step objectively and in detail.

[0070] Specifically, the resolution status includes: Resolved: the customer service response directly solved the customer's problem, or provided a clear solution, and the user did not show confusion about the unresolved problem or made further inquiries in the subsequent conversation; Unresolved: the customer service response failed to solve the customer's problem, or the solution provided was not feasible and required further follow-up processing (such as: directing the user to the information service center, information technology personnel to handle on-site, arranging engineer services, waiting for other personnel to contact the user, or issuing an order, etc.); Unclear: the content of the conversation is limited or the user does not give further feedback, and it is impossible to determine whether the problem has been resolved (such as: the user has not responded to the customer service's solution for a long time, said that it will try the suggestions but there is no follow-up feedback, or the customer service recommends that the user consult other services).

[0071] Specifically, provide the resolution: indicate whether the issue was resolved, any challenges or difficulties encountered during the resolution process, and if the issue was not resolved or was unclear, describe the plans and methods for further follow-up.

[0072] Specifically, the service evaluation may include a satisfaction score of a first object (such as a customer) and a service attitude score of a second object (such as a customer service).

[0073] For example, evaluating customer service satisfaction from the customer's perspective includes scoring customer satisfaction, including problem solving rate and accuracy, service attitude, communication effectiveness, etc. The customer satisfaction score is used to reflect the customer's satisfaction with the service process. Specifically, the better the performance in problem solving rate and accuracy, service attitude, communication effectiveness, etc., the higher the customer satisfaction score will be after weighted calculation according to a scientific and reasonable ratio. Specifically, if the customer's tone is positive, the problem is solved quickly and accurately, and the communication process is smooth and unobstructed, the customer satisfaction score is high; if the customer's tone is strongly negative, the problem is not solved at all or there is a serious communication barrier, the customer satisfaction score is low.

[0074] For example, the service attitude is scored based on the customer service's performance in the service, including politeness, professionalism, proactivity, patience, etc. The service attitude score is used to reflect the overall performance of the customer service in the service process. Specifically, the better the customer service's performance in terms of politeness, professionalism, proactivity, patience, etc., the higher the service attitude score will be after weighted calculation according to a scientific and reasonable ratio. If the customer service is always polite, respectful, professional, proactive and patient, and responds quickly, then the service attitude score is high; if the customer service has almost no politeness and professionalism, responds very slowly, and cannot communicate effectively with customers, then the service attitude score is low.

[0075] visible, Figure 2 The prompt text shown gives the problem definition and design of conversation quality inspection, guiding the model to automatically perform quality inspection on the conversation between the customer and the manual customer service, identify the problem, classify the phenomenon, judge the processing status, and give corresponding quality feedback. The prompt text can be adapted to scenarios such as intelligent customer service, customer feedback management, and conversation analysis, making the model more general and scalable.

[0076] Suppose that in a customer service conversation on an e-commerce platform, a user reports that the goods they received have quality problems. The following is an example of constructing a target prompt text: Problem information: The goods purchased by the user have quality problems, such as damage, color discrepancy, etc. Problem status: The user has reported the problem to the customer service, and the customer service is verifying the situation. Solution provided: The customer service proposes a solution of return, refund or exchange, and informs the user of the specific operation steps. Solution result: The user chooses to return and refund, the goods have been returned, and the refund has been received. Service evaluation: The user is satisfied with the customer service's response speed and solution, but is dissatisfied with the quality of the goods. Through such a target prompt text, you can clearly understand the key information in the conversation, providing strong support for subsequent quality inspection and analysis.

[0077] In this way, by including key information from multiple dimensions, the target prompt text can fully reflect all aspects of the service conversation and provide rich data support for subsequent quality inspection and analysis. The structured target prompt text makes the information clearer and more organized, making it easier for quality inspectors to quickly understand the content of the conversation and improve quality inspection efficiency. By collecting information such as service evaluation and problem solving results, companies can promptly discover deficiencies and problems in the service process and provide a strong basis for service improvement.

[0078] In some embodiments, the teacher model generates preliminary analysis results of the service conversation content based on the target prompt text, including: using the teacher model to understand the service conversation content to obtain the specific content, context, and respective intentions and concerns of the first object and the second object of the service conversation content. Based on the specific content, context, and respective intentions and concerns of the first object and the second object of the service conversation content, generate preliminary analysis results of the service conversation content according to the target prompt text.

[0079] In the disclosed embodiment, the service conversation content includes the conversation between the customer and the customer service, including the customer's questions, needs and the customer service's responses, suggestions, etc.

[0080] In the embodiments of the present disclosure, the first object usually refers to the customer, who is the recipient of the service; the second object usually refers to the customer service, who is the provider of the service.

[0081] In the disclosed embodiments, intention refers to an action goal or demand of an object in a conversation; focus refers to an aspect that an object in a conversation pays attention to or emphasizes.

[0082] In some embodiments, the teacher model first parses the service conversation content and extracts key information in the conversation, such as specific events, time, location, goods or services, etc. Based on the extracted specific content, the teacher model constructs the context of the conversation, understands the background of the conversation, historical exchanges, and the stage of the current conversation. The teacher model further analyzes the intentions and concerns of the first object (such as the customer) and the second object (such as the customer service) in the conversation. This includes identifying the customer's actual needs, emotional state, and the customer service's response strategy, service attitude, etc.

[0083] In some implementations, based on understanding the service conversation content, the teacher model matches the key information in the conversation with the target prompt text to determine whether the conversation content meets the preset quality inspection standards or business specifications. The teacher model analyzes the conversation content from multiple dimensions such as problem information, problem status, provided solutions, solution results, and service evaluation according to the requirements of the target prompt text.

[0084] In some embodiments, the teacher model integrates the analyzed information into preliminary analysis results according to the target prompt text, including evaluation results of the compliance of the conversation, customer satisfaction, service efficiency, etc.

[0085] Suppose in a customer service conversation on an e-commerce platform, a customer reports to the customer service that the purchased goods were not delivered on time. The following is an example of the preliminary analysis results generated by the teacher model:

[0086] Service conversation content understanding includes:

[0087] 1. Specific content: The goods purchased by the customer were not delivered on time, and the customer asked about the delivery time and reason.

[0088] 2. Context: The customer has urged for shipment several times before, and customer service said it is being processed.

[0089] 3. Intent and focus: Customers want to receive the goods as soon as possible and pay attention to the delivery time and reason; customer service wants to solve customers' problems and pays attention to how to comfort customers and provide solutions.

[0090] Preliminary analysis results include:

[0091] 1. Problem information: The goods purchased by the customer were not delivered on time.

[0092] 2. Problem status: The issue has been reported to customer service and is being processed.

[0093] 3. Solution provided: Customer service promised to speed up delivery and provided certain compensation measures.

[0094] 4. Resolution: Not yet clear, need to wait for follow-up.

[0095] 5. Service evaluation: The customer expressed dissatisfaction with the delayed delivery and expressed average satisfaction with the customer service’s response speed and attitude.

[0096] Through such preliminary analysis results, we can clearly understand the key information and problems in the conversation, providing strong support for subsequent service improvements and quality inspection work.

[0097] In this way, by deeply understanding the specific content and context of the service conversation, as well as the intentions and concerns of the first and second objects, the teacher model can more accurately judge the quality and problems of the conversation. The teacher model analyzes the conversation content from multiple dimensions, ensuring the comprehensiveness and depth of the analysis, which helps to discover potential service problems and improvement points. The teacher model can quickly generate preliminary analysis results, reduce the workload of manual quality inspection, improve the efficiency and accuracy of quality inspection, and provide accurate data support for the subsequent learning and optimization of the student model, which helps to improve the efficiency and accuracy of the student model's quality inspection.

[0098] In some embodiments, a preliminary analysis result of the service dialogue content is generated according to the target prompt text, including: through chain reasoning, guiding the teacher model to gradually infer the logic behind the problem according to the target prompt text: combining the logic behind the problem, and generating a preliminary analysis result of the service dialogue content according to the target prompt text.

[0099] In the disclosed embodiments, chain reasoning is a method of gradually reasoning out the logic behind things. By identifying the cause-effect relationship, time sequence and other relationships between things, a logical chain is constructed to deeply understand the essence of things.

[0100] In some embodiments, a chain reasoning method is used to gradually reason out the logic behind the problem, and generate analysis results based on these logics, including: the teacher model first understands the content of the service conversation and extracts key information in the conversation, such as the identities of the two parties in the conversation, the topic of the conversation, the questions or needs raised, etc. Based on the extracted key information, the teacher model gradually analyzes the logical relationship in the conversation content through chain reasoning. The logical relationship includes identifying the root cause of the problem, the impact of the problem on both parties, the attitudes and reactions of both parties to the problem, etc. After identifying the logic of the problem, the teacher model constructs these logics into a logical chain in chronological order or causal relationship to form a comprehensive and in-depth understanding of the problem.

[0101] In some embodiments, the preliminary analysis results are generated in combination with logic, including: after constructing the logic chain, the teacher model matches the logic information in the conversation with the target prompt text to determine which logic information meets the requirements of the target prompt text. Based on the matching results, the teacher model generates preliminary analysis results of the service conversation content according to the format and requirements of the target prompt text. For example, the preliminary analysis results include compliance assessment of the conversation, judgment of the nature of the problem, division of responsibilities between the two parties, etc. After generating the preliminary analysis results, the teacher model can conduct further review and correction to ensure the accuracy and completeness of the analysis results.

[0102] In this way, through chain reasoning, the teacher model can gradually delve into the essence of the problem and reveal the logical relationship behind the problem, thereby improving the depth and accuracy of the analysis. Building a logical chain helps the teacher model to more clearly show the ins and outs of the problem and enhance the logic and persuasiveness of the analysis results. Chain reasoning can guide the teacher model to analyze more orderly, avoid the interference of invalid information, and improve the efficiency of analysis.

[0103] In some embodiments, the student model learns and optimizes according to the preliminary analysis results, including: in the first stage, the student model is trained by imitating the preliminary analysis results output by the teacher model to learn to deal with various problems and scenarios in the conversation; in the second stage, the student model is optimized for the target conversation phenomenon by introducing sample selection and feedback enhancement mechanisms. In the third stage, the student model performs self-iteration and optimization based on its own data accumulation and feedback during the conversation process to achieve comprehensive autonomous evolution of the student model.

[0104] In the disclosed embodiments, imitation learning is a machine learning method that learns the ability to handle specific tasks by imitating the behavior of an expert or teacher model.

[0105] In the disclosed embodiment, feedback enhancement is a machine learning method that improves the training effect of the model by introducing additional supervisory information (such as scoring, labeling, etc.).

[0106] In the disclosed embodiments, self-iteration refers to the process in which the model continuously optimizes and upgrades itself based on new data and feedback in actual applications.

[0107] In the disclosed embodiment, the first stage is the imitation learning stage.

[0108] In some embodiments, the student model is trained by imitating the preliminary analysis results output by the teacher model to learn to handle various problems and scenarios in the conversation, including: collecting a large amount of service conversation content and its corresponding preliminary analysis results of the teacher model to form a training data set; the student model takes the preliminary analysis results of the teacher model as the learning goal, and imitates the processing logic and output results of the teacher model through supervised learning; by comparing the output results of the student model with the preliminary analysis results of the teacher model, the imitation learning effect of the student model is evaluated, and necessary adjustments and optimizations are made.

[0109] In the disclosed embodiment, the second stage is the optimization and improvement stage.

[0110] In some embodiments, in the second stage, the student model is optimized for the target dialogue phenomenon by introducing a sample selection and feedback enhancement mechanism, including: screening out high-quality samples related to the target dialogue phenomenon in the training data set for further optimizing the student model; introducing a manual or automatic feedback mechanism to score or annotate the output results of the student model to provide additional supervision information; based on the screened samples and feedback information, further training and optimizing the student model to improve its ability to handle specific dialogue phenomena.

[0111] In the disclosed embodiment, the third stage is the self-iteration stage.

[0112] In some embodiments, in the third stage, the student model performs self-iteration and optimization based on its own data accumulation and feedback during the conversation process to achieve comprehensive autonomous evolution of the student model, including: in actual applications, the student model continuously collects new conversation data and user feedback to form its own data accumulation; regularly evaluates the student model, analyzes its performance in actual applications, and identifies existing problems and deficiencies; based on the evaluation results and new data accumulation, the student model performs self-iteration and optimization, including adjusting model parameters, updating the knowledge base, etc., to achieve comprehensive autonomous evolution.

[0113] Figure 3 The evolution process diagram of the service dialogue quality inspection model is shown in Figure 1. Figure 3 As shown in the figure, the evolution process is divided into L0, L1, L2 and L3 stages. Among them, the L0 stage is the instruction optimization stage, which is used to optimize the target prompt text. By optimizing Prompt, the teacher model can achieve the theoretical optimal performance and provide a stable baseline effect; no labeling is required at this stage. Among them, the L1 stage (i.e. the first stage) is the imitation learning stage, and the student model is trained by imitating the preliminary analysis results output by the teacher model to learn to deal with various problems and scenarios in the dialogue. The output of the teacher model is learned with a smaller scale and more responsive student model to achieve more than 70% of the teacher model performance; no labeling is required at this stage. Among them, the L2 stage (i.e. the second stage) is the optimization and improvement stage, and the student model is optimized for the target dialogue phenomenon by introducing sample selection and feedback enhancement mechanisms. Through the second stage, the marginal effect of the imitation stage can be overcome, and the student model effect can be continuously improved to achieve 100% of the teacher model performance; no labeling is required at this stage. Among them, the L3 stage (i.e. the third stage) is the self-iteration stage, and the student model performs self-iteration and optimization based on its own data accumulation and feedback during the dialogue process to achieve comprehensive autonomous evolution of the student model. Through the third stage, the model effect can be continuously improved, and the model performance can be continuously improved through the method of model self-evolution; this stage may require one-time labeling, such as manually correcting which results output by the student model are good and which are bad.

[0114] Through the above four stages, the model evolution mechanism is adopted to enable the service dialogue quality inspection system to automatically adapt to the changes in different customer needs, significantly improving the accuracy and efficiency of dialogue quality inspection.

[0115] In this way, through imitation learning and optimization improvement, the student model can quickly master the ability to handle various problems and scenarios in the conversation and improve processing efficiency. By introducing sample selection and feedback enhancement mechanisms, the student model can optimize for specific conversation phenomena and enhance adaptability. Based on its own data accumulation and feedback mechanism, the student model can continuously iterate and optimize itself, achieve continuous evolution, and improve the quality inspection effect of the model.

[0116] In some embodiments, the student model is optimized for the target dialogue phenomenon by introducing sample selection and feedback enhancement mechanisms, including: gradually improving the student model's ability to recognize and process the target dialogue phenomenon through multiple iterative training on the selected samples; after each iteration, adjusting the parameters of the student model according to the rewards and penalties provided by the feedback enhancement mechanism.

[0117] In some embodiments, sample selection includes: screening out high-quality samples that are closely related to the target dialogue phenomenon from a large amount of dialogue data. These samples should be representative and fully reflect the characteristics of the target dialogue phenomenon. The screened samples are labeled to clarify their categories or processing results so as to provide supervision information during the training process. As the training progresses, new samples are continuously introduced to replace the old samples that have been fully learned to maintain the diversity and timeliness of the training data.

[0118] In some embodiments, a reward and penalty mechanism is designed to evaluate the performance of the student model in processing the target dialogue phenomenon. When the model correctly identifies and processes the target dialogue phenomenon, a reward is given; when the model performs poorly, a penalty is given. The reward and penalty signals can be based on a variety of factors, such as the fluency of the dialogue, user satisfaction, the accuracy of the processing results, etc. These signals can be obtained through manual annotation, user feedback, or an automatic evaluation system. According to the reward and penalty signals, the parameters of the student model are adjusted to optimize its ability to process the target dialogue phenomenon, which can be achieved through algorithms such as gradient descent and reinforcement learning.

[0119] In some embodiments, a pre-trained model or randomly initialized parameters is used as the starting point of the student model. Multiple iterations of training are performed on the selected samples. In each iteration, the student model is trained using the current sample set, and the model parameters are adjusted according to the reward and penalty signals provided by the feedback enhancement mechanism. After each iteration, the performance of the student model is evaluated using the validation set. If the performance improves, the iteration continues; if the performance decreases or stabilizes, consider adjusting the training strategy or stopping the training.

[0120] Suppose that in an intelligent customer service system, the student model needs to optimize the ability to recognize and process the dialogue phenomenon of "user consultation on product functions". The following is an example of specific implementation and effect: A large number of dialogue samples related to user consultation on product functions are screened from the dialogue log and annotated. These samples include users asking about product functions and customer service answering user questions. Design a feedback enhancement mechanism based on user satisfaction. When the user is satisfied with the processing result of product function consultation, the model is rewarded; when the user is dissatisfied, the model is punished. Perform multiple iterative training on the selected samples. In each iteration, the student model is trained using the current sample set, and the model parameters are adjusted according to user satisfaction feedback. After multiple iterative training, the student model's ability to recognize and process user consultation on product functions has been significantly improved. In actual applications, user satisfaction has been significantly improved and customer service efficiency has also been improved.

[0121] In this way, through sample selection and feedback enhancement mechanisms, the student model can more accurately identify the target dialogue phenomenon and improve the accuracy of processing. By introducing diverse samples and feedback signals, the student model can better adapt to different scenarios and user needs and enhance robustness. Through iterative training and parameter adjustment, the student model can converge to the optimal solution faster, accelerating the training process.

[0122] In some embodiments, the method further includes: evaluating the performance of the student model at the end of each stage, and deciding whether to proceed to the next stage based on the evaluation results.

[0123] In order to ensure that the student model can steadily improve in each optimization stage and smoothly enter the next stage after reaching the predetermined goal, the performance of the student model is evaluated at the end of each stage.

[0124] In some implementations, developing evaluation criteria includes: developing clear evaluation indicators such as accuracy, recall, precision, user satisfaction, etc. according to the characteristics of the target dialogue phenomenon; setting reasonable thresholds for each evaluation indicator as a basis for determining whether to enter the next stage.

[0125] In some embodiments, the evaluation process includes: at the end of each stage, preparing an independent validation data set for evaluating the performance of the student model; using the validation data set to evaluate the student model and calculate the value of each evaluation indicator; comparing the evaluation results with the set threshold value to analyze whether the performance of the student model meets the requirements.

[0126] In some embodiments, the evaluation results and analysis are used to determine whether to proceed to the next stage. If the performance is up to standard, the next stage is entered; if the performance is not up to standard, the training strategy is adjusted or training data is added, and the current stage of training is repeated.

[0127] In some implementations, the evaluation results are fed back to the model developers in a timely manner to understand the improvement of model performance and existing problems. Based on the evaluation results and feedback, the training strategy, sample selection, feedback enhancement mechanism, etc. are adjusted to improve model performance.

[0128] Suppose in an intelligent customer service system, the student model needs to optimize the ability to identify and process the dialogue phenomenon of "user consultation product function". The following is an example of specific implementation and effect:

[0129] Evaluation standard formulation: Accuracy, recall rate and user satisfaction are formulated as evaluation indicators, and reasonable thresholds are set, such as accuracy not less than 85%, recall rate not less than 80%, and user satisfaction not less than 4.5 points (out of 5 points).

[0130] Evaluation process: At the end of each stage, the student model is evaluated using an independent validation dataset. After the first stage, the evaluation results showed an accuracy of 82%, a recall of 78%, and a user satisfaction of 4.3 points, which did not meet the set threshold.

[0131] Feedback and adjustment: Feedback the evaluation results to the model developers to analyze the reasons why the model performance did not meet the standards. After analysis, it was found that the model had difficulty in identifying complex user questions. Therefore, the training strategy was adjusted to increase the number of samples of complex user questions, and the feedback enhancement mechanism was optimized to improve the model's ability to handle complex questions.

[0132] Retraining and evaluation: Retraining was performed under the adjusted training strategy and evaluated again at the end of the second phase. The evaluation results showed that the accuracy was 87%, the recall was 83%, and the user satisfaction was 4.7 points, all of which met or exceeded the set thresholds. Therefore, it was decided to enter the next phase to continue optimizing the performance of the student model.

[0133] In this way, through periodic evaluation, problems in the model can be discovered in time to avoid the accumulation of problems and the degradation of model performance. Evaluation at the end of each stage can ensure that the model enters the next stage after reaching the predetermined goal, avoid unnecessary repeated training, and improve training efficiency. Adjusting the training strategy based on the evaluation results can reasonably allocate computing resources and time and improve resource utilization efficiency.

[0134] In some embodiments, the student model learns and optimizes according to the preliminary analysis results, including: determining a target reinforcement learning method that matches the service conversation content from multiple candidate reinforcement learning methods; the student model learns and optimizes according to the preliminary analysis results based on the target reinforcement learning method.

[0135] In the disclosed embodiment, the reinforcement learning method is a machine learning algorithm that learns the optimal behavior strategy through the interaction between the intelligent agent and the environment. In the dialogue system, the reinforcement learning method can be used to optimize the dialogue strategy and improve the dialogue quality.

[0136] In the disclosed embodiment, the target reinforcement learning method is to determine the method that best matches the service dialogue content from multiple candidate reinforcement learning methods, which is used to train and optimize the student model.

[0137] In the disclosed embodiment, the preliminary analysis results are the results obtained after preliminary processing and analysis of the conversation content, including user intention, conversation status, historical information, etc. These results will be used as input data for training and optimizing the student model.

[0138] In some embodiments, determining the target reinforcement learning method includes: screening out candidate methods related to the service conversation content from an existing reinforcement learning method library. These candidate methods should be able to process key information in the conversation, such as user intent, conversation status, historical information, etc. Based on the preliminary analysis results, the applicability of each candidate method is evaluated. This includes considering the performance, stability, computational complexity, and suitability of the method for processing the current conversation content. Through comparative analysis, the target reinforcement learning method that best matches the target conversation content is determined. According to the characteristics of the target reinforcement learning method, corresponding parameters are set. These parameters may include learning rate, discount factor, exploration rate, etc., which will affect the learning speed and effect of the model.

[0139] In some embodiments, the learning and optimization of the student model includes: using the preliminary analysis results as input data, including conversation content, user intent, conversation status, etc. These data will be used to train and optimize the student model. Based on the target reinforcement learning method, the student model is trained using the input data. During the training process, the model will generate responses based on the content of the conversation and adjust its strategy based on reward signals (such as user satisfaction, conversation success rate, etc.). During the training process, the performance of the student model is continuously evaluated, and the strategy is adjusted based on the evaluation results, such as adjusting parameters, improving the model structure, or introducing new features. Through multiple iterative training, the performance of the student model is gradually optimized. After each iteration, the model is evaluated using new data, and a decision is made based on the evaluation results whether to continue training or enter the next stage.

[0140] Suppose that in an intelligent customer service system, the student model needs to optimize the processing of the conversation content of "user consultation product function". First, candidate methods related to the service conversation content are screened from the reinforcement learning method library, such as Q-learning, Deep Q-Network (DQN), etc. Then, the applicability of each candidate method is evaluated based on the preliminary analysis results (such as user intent, conversation state, etc.). Finally, DQN is determined as the target reinforcement learning method because it can handle high-dimensional state space and has good stability and performance. Using the preliminary analysis results as input data, the student model is trained based on the DQN method. During the training process, the model generates responses based on the conversation content and adjusts its strategy based on reward signals such as user satisfaction. Through multiple iterative training, the performance of the student model is gradually optimized. Finally, the student model can more accurately identify user intent and conversation state, and improve the accuracy and fluency of the conversation.

[0141] In this way, by introducing reinforcement learning methods, the student model can more accurately identify user intentions and dialogue status, thereby improving the accuracy and fluency of dialogue. The matching and parameter setting of the target reinforcement learning method enable the student model to better adapt to different dialogue scenarios and user needs, and improve the generalization ability of the model. Through iterative training and strategy optimization, the student model can learn effective dialogue strategies more quickly and improve learning efficiency.

[0142] Figure 4 An example schematic diagram of a service dialogue quality inspection model is shown, Figure 4 As shown, Figure 4 The task description (i.e., target prompt text) and dialogue examples in are input to the teacher model. In the L1 stage, the output of the teacher model is as follows Figure 4 As shown in 401, after the student model is trained by imitating the content of 401, the output of the student model is as follows Figure 4 As shown in 402, it can be seen that the output of the student model is missing fields compared with the output of the teacher model. In the L2 stage, the output of the teacher model is as follows Figure 4 As shown in 403, the student model introduces sample selection and feedback enhancement mechanisms. When dealing with complex or rare dialogue phenomena, the student model can rely on the feedback of the teacher model for targeted optimization, thereby further improving the performance. The output of the student model is as follows Figure 4 As shown in 404, it can be seen that the output of the student model is not missing fields compared with the output of the teacher model, but the answer is inaccurate. In the L3 stage, the student model no longer relies on the teacher model, but optimizes itself through its own data and dialogue feedback to achieve completely autonomous evolution of the model. The output of the current version of the student model is as follows Figure 4 As shown in 405, the output of the next version of the student model is as follows Figure 4As shown in 406, it can be seen that the output fields of the next version of the student model are complete and the information is accurate.

[0143] This disclosure introduces the concept of "model evolution", combines the student model with the teacher model, and gradually improves the quality inspection effect through an iterative optimization process, thereby realizing the self-optimization and evolution of the dialogue quality inspection system. Through the dynamic evolution mechanism, the student model can perform well in complex dialogue scenarios, gradually surpassing the performance of the teacher model, and solving the technical bottleneck that the existing quality inspection system cannot adapt to diversified needs. The gradual evolution process of the model can reduce manual intervention and improve the automation level and work efficiency of the system.

[0144] The algorithms and models disclosed in this disclosure can be applied to multiple fields, and are particularly suitable for the following scenarios: 1. Intelligent customer service: It can be used to quality inspect the quality of conversations in the intelligent customer service system, helping companies to better understand customer service performance and customer feedback. 2. Customer satisfaction management: Through conversation analysis, the system can promptly discover and address key factors that affect customer satisfaction and improve customer experience. 3. Conversation analysis: It can be used to analyze the content of conversations between customers and customer service, and provide companies with targeted service improvement suggestions. 4. Multi-round conversation quality inspection: It is particularly suitable for problem identification and processing in complex multi-round conversations, and can help customer service quickly respond to complex customer needs.

[0145] The disclosed embodiment provides a service dialogue quality inspection method based on a large model. Figure 5 : is a flowchart of a service dialogue quality inspection method based on a large model according to an embodiment of the present disclosure. The service dialogue quality inspection method can be applied to electronic devices, including but not limited to fixed devices and / or mobile devices. For example, fixed devices include but are not limited to servers, and servers can be cloud servers or ordinary servers. For example, mobile devices include but are not limited to: mobile phones, tablet computers, vehicle-mounted devices, personal computers, etc. In some possible implementations, the service dialogue quality inspection method can also be implemented by a processor calling computer-readable instructions stored in a memory. Figure 5 As shown, the service dialogue quality inspection method includes:

[0146] S501: Obtaining the content of the service dialogue to be inspected;

[0147] S502: Input the service dialogue content to be quality inspected into the service dialogue quality inspection model, and obtain the service dialogue quality inspection result output by the service dialogue quality inspection model according to the target prompt text.

[0148] In the disclosed embodiment, the target prompt text is a text used to guide the model to output the quality inspection results in the service dialogue quality inspection. It may contain specific rules, standards or requirements, and the model will evaluate the dialogue content based on these prompt texts.

[0149] In the disclosed embodiment, the service dialogue quality inspection model is used to automatically process and analyze the service dialogue content and output the quality inspection results.

[0150] In some implementations, the content of the service dialogue to be inspected may be sourced from call records of a customer service center, chat records of an online customer service system, user feedback on a social media platform, and the like.

[0151] In some embodiments, the acquired conversation content is preprocessed, including removing irrelevant information (such as noise, advertisements, etc.), standardizing text format (such as unifying capitalization, punctuation, etc.), and natural language processing steps such as word segmentation and stop word removal.

[0152] In some embodiments, the preprocessed conversation content is used as input and transmitted to the service conversation quality inspection model through an application programming interface (API) interface or other data transmission methods. After receiving the input, the service conversation quality inspection model will parse and evaluate the conversation content according to the target prompt text, and finally output the quality inspection results.

[0153] In some embodiments, the quality inspection results generally include Figure 2 The content of the prompt text displayed may also include evaluation results on the compliance of the conversation content, service quality, customer satisfaction, etc., as well as specific violation information or improvement suggestions.

[0154] Suppose a company has a customer service center that generates a large number of call records every day. In order to improve the efficiency of quality inspection, the company decided to adopt the service conversation quality inspection method. The company randomly selected 1,000 conversations from the call records of the customer service center as the content to be inspected, and transmitted the pre-processed conversation content to the model through the API interface. After receiving the input, the model evaluated each conversation and output the quality inspection results. The results showed that 30 conversations had service quality problems, such as poor customer service attitude and inaccurate answers. According to the quality inspection results, the company took corresponding improvement measures in a timely manner, such as strengthening customer service training and optimizing service processes, thereby improving service quality and customer satisfaction.

[0155] In this way, the automated service conversation quality inspection model can greatly improve the efficiency of quality inspection and reduce the time and cost of manual quality inspection. After training and tuning with a large amount of data, the model can accurately identify key information and potential problems in the conversation and improve the accuracy of quality inspection. Through the feedback of quality inspection results, enterprises can promptly identify problems in the service and take corresponding improvement measures, thereby optimizing service quality and improving customer satisfaction.

[0156] In some embodiments, the service dialogue content to be inspected includes at least one of the following: content in text form, content in voice form.

[0157] In the disclosed embodiment, the content in text form is the service conversation content presented in text form, such as emails, online customer service chat records, etc.

[0158] In the disclosed embodiment, the content in voice form is service conversation content presented in voice form, such as telephone call records, voice messages, etc.

[0159] In some embodiments, the service dialogue content in text form is collected from various channels (such as online customer service systems, emails, social media, etc.). The text is cleaned, irrelevant characters (such as tags, special symbols, etc.) are removed, and natural language processing operations such as word segmentation and part-of-speech tagging are performed. The pre-processed text content is input into the trained service dialogue quality inspection model, and the model will evaluate the quality of the text according to preset rules or algorithms.

[0160] In some implementations, speech recognition technology is used to convert speech content into text. This step may require the use of a professional speech recognition engine or tool. The converted text is preprocessed in the same way as the text content. The preprocessed text content is input into the service dialogue quality inspection model for quality assessment.

[0161] In this way, through the automated quality inspection process, the time and cost of manual quality inspection can be greatly reduced, and the efficiency of quality inspection can be improved. With the help of advanced natural language processing technology and machine learning algorithms, the service dialogue quality inspection model can accurately identify problems in the dialogue and improve the accuracy of quality inspection. This method can process service dialogue content in text and voice at the same time, meeting the quality inspection needs in different scenarios.

[0162] It should be understood that Figures 2 to 4 The schematic diagram shown is only exemplary and not restrictive, and it is scalable, and those skilled in the art can Figures 2 to 4 Various obvious changes and / or substitutions can be made to the examples, and the resulting technical solutions still fall within the scope of the disclosure of the embodiments of the present disclosure.

[0163] The present disclosure provides a training device for a service dialogue quality inspection model. Figure 6As shown, the device may include: including: a first acquisition module 601, used to acquire the target prompt text; a first control module 602, used to input the service dialogue content and the target prompt text into the teacher model, so that the teacher model generates a preliminary analysis result of the service dialogue content based on the target prompt text; a second control module 603, used to input the preliminary analysis result into the student model, so that the student model learns and optimizes according to the preliminary analysis result; a determination module 604, used to determine the student model after learning and optimization as a service dialogue quality inspection model, and the service dialogue quality inspection model is used to generate a service dialogue quality inspection result based on the target prompt text.

[0164] In some embodiments, the first acquisition module 601 includes: an acquisition submodule for acquiring an initial prompt text; and an adjustment submodule for optimizing and adjusting the initial prompt text according to the needs of the service industry to obtain a target prompt text.

[0165] In some embodiments, the target prompt text includes key information of multiple dimensions, and the multiple dimensions include: problem information, problem status, provided solutions, solution results, and service evaluation.

[0166] In some embodiments, the first control module 602 includes: an understanding sub-module, which is used to use the teacher model to understand the service dialogue content, and obtain the specific content, context, and respective intentions and concerns of the first object and the second object of the service dialogue content; the first control sub-module, which is used to generate preliminary analysis results of the service dialogue content according to the target prompt text based on the specific content, context, and respective intentions and concerns of the first object and the second object of the service dialogue content.

[0167] In some embodiments, the first control submodule 602 is used to: guide the teacher model to gradually infer the logic behind the question according to the target prompt text through chain reasoning; combine the logic behind the question and generate preliminary analysis results of the service dialogue content according to the target prompt text.

[0168] In some embodiments, the second control module 603 includes: a second control submodule, which is used for, in the first stage, the student model is trained by imitating the preliminary analysis results output by the teacher model to learn to deal with various problems and scenarios in the conversation; a third control submodule, which is used for, in the second stage, the student model is optimized for the target conversation phenomenon by introducing sample selection and feedback enhancement mechanisms; a fourth control submodule, which is used for, in the third stage, the student model performs self-iteration and optimization based on its own data accumulation and feedback during the conversation process, so as to achieve comprehensive and autonomous evolution of the student model.

[0169] In some embodiments, the third control submodule is used to: on the selected samples, through multiple iterative training, gradually improve the student model's ability to recognize and process target dialogue phenomena; after each iteration, adjust the parameters of the student model according to the rewards and penalties provided by the feedback enhancement mechanism.

[0170] In some embodiments, the second control module 603 further includes: an evaluation submodule, used to evaluate the performance of the student model at the end of each stage; and a decision submodule, used to decide whether to enter the next stage based on the evaluation results.

[0171] In some embodiments, the second control module 603 includes: a determination submodule, used to determine a target reinforcement learning method that matches the service dialogue content from multiple candidate reinforcement learning methods; and a fifth control submodule, used for the student model to learn and optimize based on the target reinforcement learning method according to the preliminary analysis results.

[0172] For the description of specific functions and examples of each module and submodule of the device in the embodiment of the present disclosure, reference can be made to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0173] The training device of the service dialogue quality inspection model of the embodiment of the present disclosure can enhance the generalization ability of the model and improve the efficiency and accuracy of service dialogue quality inspection.

[0174] The disclosed embodiment provides a service dialogue quality inspection device based on a large model, such as Figure 7 As shown, the device may include: a second acquisition module 701, used to obtain the service conversation content to be quality inspected; a quality inspection module 702, used to input the service conversation content to be quality inspected into a service conversation quality inspection model, and obtain the service conversation quality inspection result output by the service conversation quality inspection model based on the target prompt text; wherein the service conversation quality inspection model is obtained by training using the method described above.

[0175] In some embodiments, the service dialogue content to be inspected includes at least one of the following: content in text form, content in voice form.

[0176] For the description of specific functions and examples of each module and submodule of the device in the embodiment of the present disclosure, reference can be made to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0177] The large-model-based service dialogue quality inspection device of the disclosed embodiment can not only greatly improve the quality inspection efficiency and reduce the time and cost of manual quality inspection, but also improve the accuracy of quality inspection through the service dialogue quality inspection model.

[0178] The present disclosure provides a scenario diagram of a training method for a service dialogue quality inspection model. Figure 8 shown.

[0179] As mentioned above, the training method of the service dialogue quality inspection model provided by the embodiment of the present disclosure is applied to electronic devices. Electronic devices are intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices.

[0180] Specifically, the electronic device can perform the following operations:

[0181] Get the target prompt text;

[0182] Inputting the service dialogue content and the target prompt text into the teacher model, so that the teacher model generates a preliminary analysis result of the service dialogue content based on the target prompt text;

[0183] Inputting the preliminary analysis results into the student model so that the student model can learn and optimize according to the preliminary analysis results;

[0184] The student model after learning and optimization is determined as the service dialogue quality inspection model, and the service dialogue quality inspection model is used to generate service dialogue quality inspection results based on the target prompt text.

[0185] It should be understood that Figure 8 The scene diagram shown is only illustrative and not restrictive. Those skilled in the art can Figure 8 Various obvious changes and / or substitutions can be made to the examples, and the resulting technical solutions still fall within the scope of the disclosure of the embodiments of the present disclosure.

[0186] The present disclosure provides a scenario diagram of a service dialogue quality inspection method based on a large model, such as Fig. 9 shown.

[0187] As mentioned above, the service dialogue quality inspection method provided by the embodiment of the present disclosure is applied to electronic devices. Electronic devices are intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic devices can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices.

[0188] Specifically, the electronic device can perform the following operations:

[0189] Get the content of the service dialogue to be inspected;

[0190] The service dialogue content to be quality inspected is input into the service dialogue quality inspection model, and the service dialogue quality inspection result output by the service dialogue quality inspection model according to the target prompt text is obtained.

[0191] It should be understood that Fig. 9 The scene diagram shown is only illustrative and not restrictive. Those skilled in the art can Fig. 9 Various obvious changes and / or substitutions can be made to the examples, and the resulting technical solutions still fall within the scope of the disclosure of the embodiments of the present disclosure.

[0192] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0193] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0194] Fig.10 A schematic block diagram of an example electronic device 1000 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0195] like Fig.10 As shown, the device 1000 includes a computing unit 1001, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 1002 or a computer program loaded from a storage unit 1008 to a random access memory (RAM) 1003. In the RAM 1003, various programs and data required for the operation of the device 1000 can also be stored. The computing unit 1001, the ROM 1002, and the RAM 1003 are connected to each other via a bus 1004. An input / output (I / O) interface 1005 is also connected to the bus 1004.

[0196] A number of components in the device 1000 are connected to the I / O interface 1005, including: an input unit 1006, such as a keyboard, a mouse, etc.; an output unit 1007, such as various types of displays, speakers, etc.; a storage unit 1008, such as a disk, an optical disk, etc.; and a communication unit 1009, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1009 allows the device 1000 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0197] The computing unit 1001 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1001 include, but are not limited to, a CPU, a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 1001 performs the various methods and processes described above, such as the training method of the service dialogue quality inspection model and / or the service dialogue quality inspection method based on the large model. For example, in some embodiments, the training method of the service dialogue quality inspection model and / or the service dialogue quality inspection method based on the large model may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 1008. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 1000 via the ROM 1002 and / or the communication unit 1009. When the computer program is loaded into RAM 1003 and executed by the computing unit 1001, one or more steps of the service dialogue quality inspection model training method and / or the service dialogue quality inspection method based on the large model described above may be executed. Alternatively, in other embodiments, the computing unit 1001 may be configured to execute the service dialogue quality inspection model training method and / or the service dialogue quality inspection method based on the large model in any other appropriate manner (e.g., by means of firmware).

[0198] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0199] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0200] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0201] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0202] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: Local Area Networks (LANs), Wide Area Networks (WANs), and the Internet.

[0203] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0204] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0205] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A training method for a service dialogue quality inspection model, comprising: Get the target prompt text; Inputting the service dialogue content and the target prompt text into the teacher model, so that the teacher model generates a preliminary analysis result of the service dialogue content based on the target prompt text; Inputting the preliminary analysis results into the student model so that the student model can learn and optimize according to the preliminary analysis results; The student model after learning and optimization is determined as a service dialogue quality inspection model, and the service dialogue quality inspection model is used to generate a service dialogue quality inspection result according to the target prompt text.

2. The method according to claim 1, wherein: The obtaining of the target prompt text includes: Get the initial prompt text; According to the needs of the service industry, the initial prompt text is optimized and adjusted to obtain the target prompt text.

3. The method according to claim 1 or 2, wherein: The target prompt text includes key information of multiple dimensions, and the multiple dimensions include: problem information, problem status, provided solutions, solution results and service evaluation.

4. The method according to claim 1, wherein: The teacher model generates preliminary analysis results of the service dialogue content based on the target prompt text, including: The teacher model is used to understand the service dialogue content, and the specific content, context, and intentions and concerns of the first object and the second object of the service dialogue content are obtained. Based on the specific content and context of the service dialogue content and the respective intentions and concerns of the first object and the second object, a preliminary analysis result of the service dialogue content is generated according to the target prompt text.

5. The method according to claim 4, wherein: Generating a preliminary analysis result of the service dialogue content according to the target prompt text includes: Through chain reasoning, the teacher model is guided to gradually infer the logic behind the question according to the target prompt text; Combined with the logic behind the question, a preliminary analysis result of the service dialogue content is generated according to the target prompt text.

6. The method according to claim 1, wherein: The student model is learned and optimized according to the preliminary analysis results, including: In the first stage, the student model is trained by imitating the preliminary analysis results output by the teacher model to learn to handle various questions and scenarios in the conversation; In the second stage, the student model is optimized for the target dialogue phenomenon by introducing sample selection and feedback enhancement mechanisms; In the third stage, the student model performs self-iteration and optimization based on its own data accumulation and feedback during the dialogue process to achieve comprehensive autonomous evolution of the student model.

7. The method according to claim 6, wherein: The student model is optimized for the target dialogue phenomenon by introducing sample selection and feedback enhancement mechanisms, including: On the selected samples, through multiple iterative training, the student model's ability to recognize and process the target dialogue phenomenon is gradually improved; After each iteration, the parameters of the student model are adjusted according to the rewards and penalties provided by the feedback reinforcement mechanism.

8. The method according to claim 6, further comprising: At the end of each stage, the performance of the student model is evaluated, and whether to enter the next stage is determined based on the evaluation results.

9. The method according to claim 1, wherein: The student model is learned and optimized according to the preliminary analysis results, including: Determine a target reinforcement learning method that matches the service conversation content from multiple candidate reinforcement learning methods. The student model is based on the target reinforcement learning method and is learned and optimized according to the preliminary analysis results.

10. A service dialogue quality inspection method based on a large model, comprising: Get the content of the service dialogue to be inspected; The service conversation content to be quality inspected is input into a service conversation quality inspection model to obtain a service conversation quality inspection result output by the service conversation quality inspection model based on the target prompt text; wherein the service conversation quality inspection model is obtained by training using the method described in any one of claims 1 to 9.

11. The method according to claim 10, wherein: The content of the service dialogue to be inspected includes at least one of the following: content in text form and content in voice form.

12. A training device for a service dialogue quality inspection model, comprising: The first acquisition module is used to obtain the target prompt text. A first control module, configured to input the service dialogue content and the target prompt text into a teacher model, so that the teacher model generates a preliminary analysis result of the service dialogue content based on the target prompt text; A second control module, used for inputting the preliminary analysis result into the student model, so that the student model can learn and optimize according to the preliminary analysis result; A determination module is used to determine the student model after learning and optimization as a service dialogue quality inspection model, and the service dialogue quality inspection model is used to generate a service dialogue quality inspection result based on the target prompt text.

13. The device according to claim 12, wherein: The first acquisition module includes: Get submodule, used to get initial prompt text; The adjustment submodule is used to optimize and adjust the initial prompt text according to the needs of the service industry to obtain the target prompt text.

14. The device according to claim 12 or 13, wherein: The target prompt text includes key information of multiple dimensions, and the multiple dimensions include: problem information, problem status, provided solutions, solution results and service evaluation.

15. The device according to claim 12, wherein: The first control module comprises: The understanding submodule is used to understand the service dialogue content by using the teacher model to obtain the specific content, context, and intentions and concerns of the first object and the second object. The first control submodule is used to generate a preliminary analysis result of the service dialogue content according to the target prompt text based on the specific content and context of the service dialogue content and the intentions and concerns of the first object and the second object respectively.

16. The device according to claim 15, wherein: The first control submodule is used to: Through chain reasoning, the teacher model is guided to gradually infer the logic behind the question according to the target prompt text; Combined with the logic behind the question, a preliminary analysis result of the service dialogue content is generated according to the target prompt text.

17. The device according to claim 12, wherein: The second control module comprises: The second control submodule is used for, in the first stage, the student model is trained by imitating the preliminary analysis results output by the teacher model to learn to handle various problems and scenarios in the dialogue. The third control submodule is used to optimize the target dialogue phenomenon by introducing sample selection and feedback enhancement mechanism in the second stage. The fourth control submodule is used for, in the third stage, the student model to perform self-iteration and optimization based on its own data accumulation and feedback during the dialogue process, so as to achieve comprehensive autonomous evolution of the student model.

18. The device according to claim 17, wherein: The third control submodule is used to: On the selected samples, through multiple iterative training, the student model's ability to recognize and process the target dialogue phenomenon is gradually improved; After each iteration, the parameters of the student model are adjusted according to the rewards and penalties provided by the feedback reinforcement mechanism.

19. The device according to claim 17, wherein: The second control module further includes: An evaluation submodule, for evaluating the performance of the student model at the end of each stage; The decision submodule is used to decide whether to enter the next stage based on the evaluation results.

20. The device according to claim 12, wherein: The second control module comprises: A determination submodule, configured to determine a target reinforcement learning method that matches the service dialogue content from a plurality of candidate reinforcement learning methods; The fifth control submodule is used for the student model to learn and optimize according to the preliminary analysis results based on the target reinforcement learning method.

21. A service dialogue quality inspection device based on a large model, comprising: The second acquisition module is used to obtain the content of the service dialogue to be inspected; A quality inspection module is used to input the service dialogue content to be quality inspected into a service dialogue quality inspection model to obtain the service dialogue quality inspection result output by the service dialogue quality inspection model based on the target prompt text; wherein the service dialogue quality inspection model is obtained by training using the method described in any one of claims 1 to 9.

22. The device according to claim 21, wherein The content of the service dialogue to be inspected includes at least one of the following: content in text form and content in voice form.

23. An electronic device, comprising: at least one processor; as well as A memory communicatively coupled to at least one processor, wherein: The memory stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor to enable the at least one processor to perform the method of any one of claims 1-11.

24. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are for causing a computer to perform a method according to any one of claims 1-11.

25. A computer program product comprising a computer program stored on a storage medium, the computer program implementing the method according to any one of claims 1 to 11 when executed by a processor.

Images