Customer service session content classification method and device
By integrating the classification tree and session data, and using the Prompt project of the big model for hierarchical classification, the problems of low efficiency and poor accuracy of customer service session content classification in the existing technology are solved, and more efficient classification results are achieved.
Patent Information
- Application Number
- CN202411305734.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-18
- Publication Date
- 2024-12-31
AI Technical Summary
In the existing customer service system, customer service personnel need to manually select problem categories, resulting in increased work difficulty and reduced efficiency. The traditional content classification model is poor in effect and high in cost when facing thousands of categories.
By obtaining the classification tree and customer service session data, integrating classification prompt text at different levels, and calling preset large models for classification, generating Prompt prompt text to realize the classification of custom rules, and using hierarchical classification methods to improve accuracy.
It reduces the cost of classifying customer service session content, improves classification accuracy and effectiveness, and avoids the problem of inaccurate classification in one-time.
Smart Images
Figure CN119226512A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present application relate to the field of Internet technology, and in particular, to a method, apparatus, computer device, computer-readable storage medium, and computer program product for classifying customer service conversation content. Background Art
[0002] With the rapid development of the Internet service industry, each network platform needs a large number of customer service personnel to communicate with users to solve service and communication problems. In the existing customer service system, after handling the user's problem, the customer service personnel need to summarize the customer service conversation based on the current customer problem, that is, select a category from a large number of problem categories to mark the customer service conversation. However, the current process requires manual operation and manual screening of the problem category by the customer service personnel. Obviously, this operation method greatly increases the difficulty of the customer service personnel's work and affects their daily work efficiency.
[0003] To this end, the industry has begun to try to use various machine learning-related technologies to solve the above problems. However, since there are many categories in the customer service conversation classification summary, even thousands of them, the traditional content classification model is difficult to support the classification of thousands of categories, and the training data requirements are very high. It is necessary to prepare a huge amount of data for training, which is extremely costly. Moreover, when faced with thousands of categories, the traditional classification method is relatively ineffective because its classification method is relatively mechanical and is not based on text understanding, but on word features.
[0004] It should be noted that the above content is not necessarily prior art, nor is it intended to limit the scope of patent protection of this application. Summary of the invention
[0005] Embodiments of the present application provide a method, apparatus, computer device, computer-readable storage medium, and computer program product for classifying customer service conversation content to solve or alleviate one or more of the technical problems raised above.
[0006] One aspect of an embodiment of the present application provides a method for classifying customer service conversation content, the method comprising: Get classification tree and customer service session data; Integrating the classification trees to obtain a plurality of classification prompt texts at different levels; For each level of classification prompt text, a preset large model is called to classify the customer service conversation data to obtain a classification result for each level; The classification results of all levels are integrated to obtain the target classification result.
[0007] Optionally, for each level of classification prompt text, calling a preset large model to classify the customer service conversation data to obtain classification results for each level includes: Generate a Prompt prompt text according to the classification prompt text of each level and the customer service session data; According to the Prompt text, the preset large model is called to perform classification to obtain the classification results of each level.
[0008] Optionally, generating a Prompt prompt text according to the classification prompt text of each level and the customer service session data includes: Get the classification prompt template; The classification prompt text of each level and the customer service session data are filled in the classification prompt template to obtain the Prompt prompt text.
[0009] Optionally, the classification prompt template includes a classification rule summary area and a conversation text area; the classification prompt text of each level and the customer service conversation data are filled in the classification prompt template to obtain the Prompt prompt text, including: The classification prompt text of each level is filled into the classification rule summary area, and the customer service conversation data is filled into the conversation text area to obtain the Prompt prompt text.
[0010] Optionally, obtaining customer service session data includes: Obtaining the ended session identifier regularly according to a preset period; Acquire customer service session data corresponding to the session identifier.
[0011] Optionally, the classification tree includes multiple classification categories, each of which has a one-to-one corresponding level; the classification tree is integrated to obtain classification prompt texts of multiple different levels, including: According to the level corresponding to each classification category in the classification tree, the classification categories at the same level are integrated respectively to obtain classification prompt texts at multiple different levels.
[0012] Optionally, integrating the classification results of all levels to obtain the target classification result includes: Determine the level corresponding to the classification prompt text as the level corresponding to the classification result; According to the level corresponding to each classification result, the classification results of all levels are integrated to obtain the target classification result.
[0013] Another aspect of an embodiment of the present application provides a device for classifying customer service conversation content, the device comprising: Classification tree acquisition module, used to obtain classification trees and customer service session data; A prompt text integration module, used for integrating the classification tree to obtain a plurality of classification prompt texts at different levels; A conversation data classification module is used to call a preset large model to classify the customer service conversation data for each level of classification prompt text to obtain a classification result for each level; The classification result integration module is used to integrate the classification results of all levels to obtain the target classification result.
[0014] Another aspect of an embodiment of the present application provides a computer device, including: at least one processor; and a memory communicatively coupled to the at least one processor; Wherein: 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 so that the at least one processor can execute the above-mentioned customer service conversation content classification method.
[0015] Another aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method for classifying customer service conversation content as described above is implemented.
[0016] Another aspect of an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method for classifying customer service conversation content as described above.
[0017] The technical solution adopted in the embodiment of the present application may include the following advantages: by using the Prompt project of a large model to summarize and classify the customer service conversations, classification can be performed based on custom rules instead of training the classification model based on the traditional machine learning supervision method, thereby reducing the cost of classifying the customer service conversation content. Moreover, a hierarchical classification method is adopted in the classification process, which can avoid the problem of inaccurate classification summary caused by one-time classification, thereby improving the accuracy of customer service conversation content classification and achieving better classification effect. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The accompanying drawings exemplarily illustrate the embodiments and constitute a part of the specification, and together with the text description of the specification, are used to explain the exemplary implementation of the embodiments. The embodiments shown are for illustrative purposes only and do not limit the scope of the claims. In all drawings, the same reference numerals refer to similar but not necessarily identical elements.
[0019] Figure 1A schematic diagram schematically shows a user interface for customer service conversation classification in the prior art; Figure 2 A diagram schematically shows an operating environment of a method for classifying customer service conversation content according to Embodiment 1 of the present application; Figure 3 A flowchart schematically shows a method for classifying customer service conversation content according to Embodiment 1 of the present application; Figure 4 A schematic diagram of calling a preset large model for classification to output classification results is schematically shown; Figure 5 A block diagram schematically shows a device for classifying customer service conversation content according to Embodiment 2 of the present application; and Figure 6 The hardware architecture diagram of the computer device according to the third embodiment of the present application is schematically shown. DETAILED DESCRIPTION
[0020] In order to make the purpose, technical solutions and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application.
[0021] It should be noted that the descriptions involving "first", "second", etc. in the embodiments of the present application are only for descriptive purposes and cannot be understood as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in the field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be deemed that such combination of technical solutions does not exist and is not within the scope of protection required by this application.
[0022] In the description of the present application, it should be understood that the numerical labels before the steps do not indicate the order in which the steps are executed, but are only used to facilitate the description of the present application and to distinguish each step, and therefore should not be understood as a limitation on the present application.
[0023] First, the following terms are explained: Prompt: is an instruction, question or sentence that can be used to guide or instruct a language model to generate specific text output. It is the starting point for users to interact with language models, telling the model the user's intention and expecting the model to respond in a meaningful and relevant way.
[0024] Prompt engineering: is the practice of developing and optimizing these prompts so that language models can be effectively used in various applications. Through carefully designed prompts, large language models can be guided to better understand user intent and generate more accurate and useful responses.
[0025] Big model: also known as deep learning model, is an artificial intelligence technology that has emerged in recent years with the rapid development of computer technology and big data. The basic principle of big model is to use the structure of neural network to automatically adjust the connection weights between nodes through the training process to adapt to different tasks and data.
[0026] Secondly, in order to facilitate those skilled in the art to understand the technical solutions provided in the embodiments of the present application, the relevant technologies are described below: With the rapid development of the Internet service industry, each network platform needs a large number of customer service personnel to communicate with users to solve problems in service and communication. In the existing customer service system, after handling the user's problem, the customer service personnel need to summarize the customer service conversation based on the current customer problem, that is, select a category from a large number of problem categories to mark the customer service conversation. However, the current process requires manual operation and manual screening of the problem category by the customer service personnel. Obviously, this operation method greatly increases the difficulty of the customer service personnel's work and affects their daily work efficiency. Figure 1 , which is a schematic diagram of a user interface for customer service conversation classification in the prior art.
[0027] To this end, the industry has begun to try to use various machine learning-related technologies to solve the above problems. However, since there are many categories in the customer service conversation classification summary, even thousands of them, traditional content classification models (such as Bayesian classification based on machine learning, or SVM classification) are difficult to support the classification of thousands of categories, and the training data requirements are very high. It is necessary to prepare huge data for training, which is extremely costly. Moreover, when faced with thousands of categories, the traditional classification method is relatively ineffective because its classification method is relatively mechanical and is not based on text understanding, but on word features.
[0028] To this end, an embodiment of the present application provides a technical solution for classifying customer service conversation content. In this technical solution, a classification tree and customer service conversation data are obtained; the classification tree is integrated to obtain classification prompt texts of multiple different levels; for each level of classification prompt text, a preset large model is called to classify the customer service conversation data to obtain the classification results of each level; and the classification results of all levels are integrated to obtain the target classification results. By using the Prompt project of the large model to summarize and classify the customer service conversations, classification can be performed based on custom rules instead of training the classification model based on the traditional machine learning supervision method, thereby reducing the cost of classifying the customer service conversation content. Moreover, the classification process adopts a hierarchical classification method to classify, which can avoid the problem of inaccurate classification summary caused by one-time classification, thereby improving the accuracy of customer service conversation content classification and achieving better classification effect. See below for details.
[0029] Finally, for ease of understanding, an exemplary operating environment is provided below.
[0030] like Figure 2 As shown, the environment diagram includes a service platform 2, a network 4, and a client 6, wherein: The service platform 2 may be composed of a single or multiple computing devices. The multiple computing devices may include virtualized computing instances. Virtualized computing instances may include virtual machines, such as simulations of computer systems, operating systems, servers, etc. The computing device may load a virtual machine based on a virtual image and / or other data defining specific software (e.g., operating system, dedicated application, server) for simulation. As the demand for different types of processing services changes, different virtual machines may be loaded and / or terminated on one or more computing devices. A hypervisor may be implemented to manage the use of different virtual machines on the same computing device.
[0031] The service platform 2 may be configured to communicate with the client 6 or the like via a network 4. The network 4 includes various network devices, such as routers, switches, multiplexers, hubs, modems, bridges, repeaters, firewalls, proxy devices, and / or the like. The network 4 may include physical links, such as coaxial cable links, twisted pair cable links, optical fiber links, combinations thereof, and the like, or wireless links, such as cellular links, satellite links, Wi-Fi links, and the like.
[0032] The service platform 2 can provide storage, reading, writing, querying, deleting and other services, such as providing classification services of customer service conversation content to the client.
[0033] The client 6 may be an electronic device running an operating system such as Windows, Android™ or iOS, such as a smart phone, a tablet device, a laptop computer, a virtual reality device, a gaming device, a set-top box, a vehicle terminal, or a smart TV. Based on the above operating system, various applications may be run, such as an application for classifying customer service conversation content.
[0034] The client 6 may provide / configure a user access page for manipulating the service platform 2 or uploading objects, etc.
[0035] It should be noted that the above devices are exemplary, and the number and type of devices are adjustable in different scenarios or according to different needs.
[0036] The following uses the service platform 2 as the execution subject to introduce the technical solution of the present application through multiple embodiments. It should be noted that these embodiments can be implemented in a variety of different forms and should not be interpreted as being limited to the embodiments described here.
[0037] Embodiment 1 Figure 3 The flowchart schematically shows a method for classifying customer service conversation content according to the first embodiment of the present application.
[0038] like Figure 3 As shown, the classification method of the customer service conversation content may include steps S302 to S308, wherein: Step S302, obtaining a classification tree and customer service session data; Step S304, integrating the classification tree to obtain a plurality of classification prompt texts at different levels; Step S306, for each level of classification prompt text, calling a preset large model to classify the customer service conversation data to obtain a classification result for each level; Step S308: Integrate the classification results of all levels to obtain the target classification result.
[0039] The classification method for customer service conversation content provided in this embodiment uses the Prompt project of a large model to summarize and classify customer service conversations, so that classification can be performed based on user-defined rules instead of training a classification model based on a traditional machine learning supervision method, thereby reducing the cost of classifying the customer service conversation content. Moreover, a hierarchical classification method is adopted in the classification process, which can avoid the problem of inaccurate classification summary caused by one-time classification, thereby improving the accuracy of customer service conversation content classification and achieving better classification effect.
[0040] The following combination Figure 3 , each step in steps S302~S308 and other optional steps are described in detail.
[0041] Step S302 , obtain the classification tree and customer service session data; The classification tree is a data structure based on nodes and edges, and each node in the classification tree corresponds to a classification category. In this embodiment, the classification tree includes multiple classification categories, and each classification category has a one-to-one corresponding level.
[0042] The customer service session data is session data generated when a customer service staff communicates with a user in a customer service system. In this embodiment, the customer service session data may be a session data that has ended.
[0043] Step S304 , integrating the classification trees to obtain a plurality of classification prompt texts at different levels; In this embodiment, the classification tree can be integrated according to the hierarchical relationship between nodes in the classification tree to obtain multiple classification prompt texts of different levels. Specifically, the classification categories at the same level can be integrated together to form a classification prompt text.
[0044] Step S306 , for each level of classification prompt text, calling the preset large model to classify the customer service conversation data to obtain the classification results of each level; Among them, the preset large model is a method or technology in the field of artificial intelligence for processing large-scale data and complex models. The preset large model is built by a deep neural network, has billions or even hundreds of billions of parameters, and can handle more complex tasks and data.
[0045] In this embodiment, through hierarchical classification, for each level of classification prompt text, the preset big model is called to classify the customer service conversation data to obtain the classification results of each level. Therefore, during the classification process, the preset big model is called multiple times and the classification results of this sub-classification are output.
[0046] Step S308 , integrating the classification results of all levels to obtain the target classification result.
[0047] In this embodiment, the target classification result can be obtained by integrating the classification results of all levels according to the level corresponding to each classification result.
[0048] In an optional embodiment of the present application, step S306 may include the following steps: Generate a Prompt prompt text according to the classification prompt text of each level and the customer service conversation data; call a preset large model to perform classification according to the Prompt prompt text to obtain a classification result of each level.
[0049] In this embodiment, the classification rules can be defined based on the Prompt project to instruct the preset big model to perform classification according to the custom rules and output the classification results. The Prompt prompt text is generated according to the classification prompt text of each level and the customer service session data; and the preset big model is called according to the Prompt prompt text to perform classification to obtain the classification results of each level.
[0050] In an optional embodiment of the present application, generating a Prompt prompt text according to the classification prompt text of each level and the customer service session data includes: Obtain a classification prompt template; fill the classification prompt text of each level and the customer service session data in the classification prompt template to obtain a Prompt prompt text.
[0051] In this embodiment, a classification prompt template is pre-designed, and the classification prompt template is used to specify the format of the Prompt prompt text. The Prompt prompt text is obtained by obtaining the classification prompt template, and then filling the classification prompt text of each level and the customer service session data in the classification prompt template.
[0052] In an optional embodiment of the present application, the classification prompt template includes a classification rule summary area and a conversation text area; the classification prompt text of each level and the customer service conversation data are filled in the classification prompt template to obtain the Prompt prompt text, including: The classification prompt text of each level is filled into the classification rule summary area, and the customer service conversation data is filled into the conversation text area to obtain the Prompt prompt text.
[0053] In this embodiment, the classification prompt template includes a classification rule summary area and a conversation text area. The Prompt prompt text is obtained by filling the classification prompt text of each level into the classification rule summary area and filling the customer service conversation data into the conversation text area.
[0054] As an example, a category hint template might look like this: Here is a summary description of each: {Category 1} (category ID is {1}): contains {} {Category 2} (category ID is {2}): contains {} {Category 3} (category ID is {3}): contains {} {Category 4} (category ID is {4}): contains {} The rule is, I give you a dialogue text, you tell me the category ID, below is the dialogue text, and you directly return the category ID.
[0055] Dialogue Transcript: { } Please return the category ID The curly brackets {} in the above classification prompt template are used to obtain relevant data for filling.
[0056] As an example, the prompt text obtained after filling in the relevant data is as follows: Here is a summary description of each: E-commerce (category ID is 1): including logistics, pre-sales, after-sales, product quality, return / refund, member purchase, business cooperation, activities, sudden failure General business (category ID is 2): including refunds for minors, games, digital currency, etc. Live broadcast business (category ID is 3): including live broadcast users, live broadcast host related, live broadcast guild related, live broadcast sudden failure Invalid conversations (category ID is 4): Contains meaningless conversations or those with little content The rule is, I give you a dialogue text, you tell me the category ID, below is the dialogue text, and you directly return the category ID.
[0057] Dialogue Transcript: User: No live streaming for 180 days User: Resigned Customer Service: Hello, buddy (⁄ ⁄•⁄ω⁄•⁄ ⁄)⁄ User: Hello Customer Service: In this case, you need to go to your mobile phone Live Ji - Anchor Center - My Guild and directly initiate the withdrawal. User: OK Customer Service: Thank you for your hard work and your understanding~ Please return the category ID.
[0058] As an example, Figure 4 A schematic diagram showing a method of calling a preset large model for classification to output classification results is shown. Figure 4 In the example, the Prompt text generated above is used as the input of the preset large model. After calling the preset large model for classification, the output classification ID is 3.
[0059] In an optional embodiment of the present application, the obtaining of customer service session data includes: Obtain the ended session identifier regularly according to a preset period; and obtain the customer service session data corresponding to the session identifier.
[0060] In this embodiment, a scheduled task Job may be registered, and the scheduled task Job may be used to obtain the ended session identifier according to a preset period, and obtain the customer service session data corresponding to the session identifier.
[0061] The session identifier is a unique identifier corresponding to the session window used for communication.
[0062] The preset period is a pre-set time interval for acquiring data. For example, the preset period may be 1 hour or 2 hours, etc. In a specific implementation, it may be set according to actual needs, and the embodiment of the present application does not impose any specific restrictions on this.
[0063] In an optional embodiment of the present application, the classification tree includes multiple classification categories, each classification category has a one-to-one corresponding level; the classification tree is integrated to obtain multiple classification prompt texts of different levels, including: According to the level corresponding to each classification category in the classification tree, the classification categories at the same level are integrated respectively to obtain classification prompt texts at multiple different levels.
[0064] In this embodiment, the classification tree contains multiple classification categories, each of which has a one-to-one corresponding level. By integrating the classification categories at the same level according to the level corresponding to each classification category in the classification tree, multiple classification prompt texts at different levels are obtained. That is, the classification categories at the same level are integrated together as a classification prompt text, so that the interface of the preset large model can be called hierarchically according to the classification prompt text for classification, so as to achieve the purpose of hierarchical classification, improve the accuracy of classification, and enhance the classification effect of customer service conversation data.
[0065] In an optional embodiment of the present application, integrating the classification results of all levels to obtain the target classification result includes: The level corresponding to the classification prompt text is determined as the level corresponding to the classification result; and according to the level corresponding to each classification result, the classification results of all levels are integrated to obtain the target classification result.
[0066] In this embodiment, since classification is performed by hierarchical calling of the interface of the preset large model, the level corresponding to the classification prompt text input into the preset large model can be used as the level corresponding to the classification result output by the sub-classification, and then, according to the level corresponding to each classification result, the classification results of all levels are integrated to obtain the target classification result.
[0067] The data involved in the embodiments of the present application (including but not limited to customer service session data, etc.) are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.
[0068] Embodiment 2 Figure 5 The block diagram of the classification device for customer service conversation content according to the second embodiment of the present application is schematically shown. The device can be divided into one or more program modules, one or more program modules are stored in a storage medium, and are executed by one or more processors to complete the embodiment of the present application. The program module referred to in the embodiment of the present application refers to a series of computer program instruction segments that can complete specific functions. The following description will specifically introduce the functions of each program module in this embodiment. Figure 5 As shown, the customer service conversation content classification device 500 may include: a classification tree acquisition module 510, a prompt text integration module 520, a conversation data classification module 530 and a classification result integration module 540, wherein: A classification tree acquisition module 510 is used to acquire classification trees and customer service session data; A prompt text integration module 520 is used to integrate the classification tree to obtain a plurality of classification prompt texts at different levels; The conversation data classification module 530 is used to call a preset large model to classify the customer service conversation data for each level of classification prompt text to obtain a classification result for each level; The classification result integration module 540 is used to integrate the classification results of all levels to obtain the target classification result.
[0069] In an optional embodiment of the present application, the session data classification module 530 includes: A prompt text generation submodule, used to generate a Prompt prompt text according to the classification prompt text of each level and the customer service session data; The session data classification module is used to call the preset large model to classify according to the Prompt prompt text to obtain the classification results of each level.
[0070] In an optional embodiment of the present application, the prompt text generation submodule includes: A template acquisition unit, used to acquire a classification prompt template; The prompt text generating unit is used to fill the classification prompt text of each level and the customer service session data in the classification prompt template to obtain the Prompt prompt text.
[0071] In an optional embodiment of the present application, the classification prompt template includes a classification rule summary area and a conversation text area; the prompt text generation unit includes: The prompt text generating subunit is used to fill the classification prompt text of each level into the classification rule summary area, and fill the customer service conversation data into the conversation text area to obtain the Prompt prompt text.
[0072] In an optional embodiment of the present application, the classification tree acquisition module 510 includes: The session identifier acquisition submodule is used to obtain the ended session identifier according to a preset period; The customer service session data submodule is used to obtain the customer service session data corresponding to the session identifier.
[0073] In an optional embodiment of the present application, the classification tree includes multiple classification categories, each classification category has a one-to-one corresponding level; the prompt text integration module 520 includes: The prompt text integration submodule is used to integrate the classification categories at the same level according to the level corresponding to each classification category in the classification tree, so as to obtain a plurality of classification prompt texts at different levels.
[0074] In an optional embodiment of the present application, the classification result integration module 540 includes: A classification result level determination submodule, used to determine the level corresponding to the classification prompt text as the level corresponding to the classification result; The classification result integration submodule is used to integrate the classification results of all levels according to the level corresponding to each classification result to obtain the target classification result.
[0075] Embodiment 3 Figure 6 The schematic diagram of the hardware architecture of a computer device 10000 suitable for implementing the classification method of customer service conversation content according to the third embodiment of the present application is shown schematically. In some embodiments, the computer device 10000 may be a terminal device such as a smart phone, a wearable device, a tablet computer, a personal computer, a vehicle terminal, a game console, a virtual device, a workbench, a digital assistant, a set-top box, a robot, etc. In other embodiments, the computer device 10000 may be a rack server, a blade server, a tower server, or a cabinet server (including an independent server, or a server cluster composed of multiple servers), etc. Figure 6 As shown, the computer device 10000 includes but is not limited to: a memory 10010, a processor 10020, and a network interface 10030 that can communicate with each other through a system bus. Among them: The memory 10010 includes at least one type of computer-readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 10010 can be an internal storage module of the computer device 10000, such as a hard disk or memory of the computer device 10000. In other embodiments, the memory 10010 can also be an external storage device of the computer device 10000, such as a plug-in hard disk equipped on the computer device 10000, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. Of course, the memory 10010 can also include both the internal storage module of the computer device 10000 and its external storage device. In this embodiment, the memory 10010 is generally used to store the operating system and various application software installed on the computer device 10000, such as the program code of the customer service conversation content classification method, etc. In addition, the memory 10010 can also be used to temporarily store various data that have been output or will be output.
[0076] In some embodiments, the processor 10020 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other chips. The processor 10020 is generally used to control the overall operation of the computer device 10000, such as performing control and processing related to data interaction or communication with the computer device 10000. In this embodiment, the processor 10020 is used to run the program code stored in the memory 10010 or process data.
[0077] The network interface 10030 may include a wireless network interface or a wired network interface, and the network interface 10030 is generally used to establish a communication link between the computer device 10000 and other computer devices. For example, the network interface 10030 is used to connect the computer device 10000 to an external terminal through a network, and to establish a data transmission channel and a communication link between the computer device 10000 and the external terminal. The network may be a wireless or wired network such as an intranet, the Internet, the Global System of Mobile communication (GSM), Wideband Code Division Multiple Access (WCDMA), 4G network, 5G network, Bluetooth, Wi-Fi, etc.
[0078] It should be pointed out that Figure 6 Only a computer device having components 10010 - 10030 is shown, but it should be understood that implementation of all of the components shown is not a requirement, and more or fewer components may alternatively be implemented.
[0079] In this embodiment, the classification method of the customer service conversation content stored in the memory 10010 can also be divided into one or more program modules and executed by one or more processors (such as processor 10020) to complete the embodiment of the present application.
[0080] Embodiment 4 An embodiment of the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the method for classifying customer service conversation content in the embodiment are implemented.
[0081] In this embodiment, the computer-readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the computer-readable storage medium may be an internal storage unit of a computer device, such as a hard disk or memory of the computer device. In other embodiments, the computer-readable storage medium may also be an external storage device of a computer device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device. Of course, the computer-readable storage medium may also include both an internal storage unit of a computer device and an external storage device thereof. In this embodiment, the computer-readable storage medium is generally used to store an operating system and various application software installed on a computer device, such as the program code of the classification method of the customer service conversation content in the embodiment, etc. In addition, the computer-readable storage medium may also be used to temporarily store various types of data that have been output or are to be output.
[0082] Embodiment 5 An embodiment of the present application also provides a computer program product, including a computer program, which implements the method in the above embodiment when executed by a processor.
[0083] Obviously, those skilled in the art should understand that the modules or steps of the above-mentioned embodiments of the present application can be implemented by general-purpose computer devices, they can be concentrated on a single computer device, or distributed on a network composed of multiple computer devices, optionally, they can be implemented by executable program codes of computer devices, so that they can be stored in a storage device and executed by the computer device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.
[0084] It should be noted that the above are only preferred embodiments of the present application, and the patent protection scope of the present application is not limited thereto. Any equivalent structure or equivalent process transformation made using the contents of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.
Claims
1. A method for classifying customer service conversation content, characterized in that: The method comprises: Get classification tree and customer service session data; Integrating the classification trees to obtain a plurality of classification prompt texts at different levels; For each level of classification prompt text, a preset large model is called to classify the customer service conversation data to obtain a classification result for each level; The classification results of all levels are integrated to obtain the target classification result.
2. The method for classifying customer service conversation content according to claim 1, characterized in that: For each level of classification prompt text, calling a preset large model to classify the customer service conversation data to obtain classification results for each level includes: Generate a Prompt prompt text according to the classification prompt text of each level and the customer service session data; According to the Prompt text, the preset large model is called to perform classification to obtain the classification results of each level.
3. The method for classifying customer service conversation content according to claim 2, characterized in that: The generating of the Prompt prompt text according to the classification prompt text of each level and the customer service session data includes: Get the classification prompt template; The classification prompt text of each level and the customer service session data are filled in the classification prompt template to obtain the Prompt prompt text.
4. The method for classifying customer service conversation content according to claim 3, characterized in that: The classification prompt template includes a classification rule summary area and a conversation text area; the classification prompt text of each level and the customer service conversation data are filled in the classification prompt template to obtain the Prompt prompt text, including: The classification prompt text of each level is filled into the classification rule summary area, and the customer service conversation data is filled into the conversation text area to obtain the Prompt prompt text.
5. The method for classifying customer service conversation content according to claim 1, characterized in that: The obtaining of customer service session data includes: Obtaining the ended session identifier regularly according to a preset period; Acquire customer service session data corresponding to the session identifier.
6. The method for classifying customer service conversation content according to claim 1, characterized in that: The classification tree includes multiple classification categories, each of which has a one-to-one corresponding level; the classification tree is integrated to obtain multiple classification prompt texts of different levels, including: According to the level corresponding to each classification category in the classification tree, the classification categories at the same level are integrated respectively to obtain classification prompt texts at multiple different levels.
7. The method for classifying customer service conversation content according to claim 1, characterized in that: The integration of classification results at all levels to obtain the target classification result includes: Determine the level corresponding to the classification prompt text as the level corresponding to the classification result; According to the level corresponding to each classification result, the classification results of all levels are integrated to obtain the target classification result.
8. A classification device for customer service conversation content, characterized in that: The device comprises: Classification tree acquisition module, used to obtain classification trees and customer service session data; A prompt text integration module, used for integrating the classification tree to obtain a plurality of classification prompt texts at different levels; A conversation data classification module is used to call a preset large model to classify the customer service conversation data for each level of classification prompt text to obtain a classification result for each level; The classification result integration module is used to integrate the classification results of all levels to obtain the target classification result.
9. A computer device, characterized in that: include: at least one processor; and a memory communicatively connected to the at least one processor; wherein: 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 the method for classifying customer service conversation content according to any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and when the computer instructions are executed by a processor, the method for classifying customer service conversation content according to any one of claims 1 to 7 is implemented.
11. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for classifying customer service conversation content described in claims 1 to 7 are implemented.