User question processing method and device, electronic equipment and storage medium

Abstract

The invention discloses a user question processing method and device, electronic equipment and a storage medium, and relates to the field of computers, in particular to the technical field of artificial intelligence such as large models, deep learning, natural language processing and intelligent customer service. According to the specific implementation scheme, the customer service system firstly obtains a current to-be-processed user question and reference data associated with the user question, then performs intention recognition on the user question based on the reference data, determines a target intention of the user question, and sends the user question to the customer service system under the condition that the target intention is matched with any reference intention. According to the parameter description information associated with any reference intention, target parameters are extracted from the user question and the reference data, and finally the target parameters are sent to a server through an application programming interface (API) associated with any reference intention.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technologies, and particularly to artificial intelligence technology fields such as large models, deep learning, natural language processing, intelligent customer service, etc., and specifically relates to a method, apparatus, electronic device, and storage medium for processing user problems. Background Art

[0002] Current intelligent customer services usually process text problems, and after determining a solution, manual customer service performs operations such as manually restarting or changing virtual machines. However, this method cannot effectively integrate multimodal data, and manual operations are prone to errors, resulting in low efficiency and accuracy in processing user problems. Summary of the Invention

[0003] The present disclosure provides a method, apparatus, electronic device, and storage medium for processing user problems. The specific solutions are as follows:

[0004] According to one aspect of the present disclosure, there is provided a method for processing user problems, characterized by comprising:

[0005] Obtain the currently to-be-processed user problem and reference data associated with the user problem;

[0006] Based on the reference data, perform intent recognition on the user problem to determine the target intent of the user problem;

[0007] In the case where the target intent matches any reference intent, extract target parameters from the user problem and the reference data according to the parameter description information associated with the any reference intent;

[0008] Send the target parameters to a server through an application programming interface (API) associated with the any reference intent.

[0009] According to another aspect of the present disclosure, there is provided a method for processing user problems, characterized by comprising:

[0010] Receive target parameters sent by a customer service system through any application programming interface (API);

[0011] Determine a target virtual machine to be processed and target operation parameters according to the target parameters and the any API;

[0012] Process the target virtual machine based on the target operation parameters.

[0013] According to another aspect of the present disclosure, there is provided a user problem processing apparatus, characterized by comprising:

[0014] An acquisition module, configured to acquire a current user problem to be processed and reference data associated with the user problem;

[0015] A first determination module, configured to perform intent recognition on the user problem based on the reference data to determine the target intent of the user problem;

[0016] An extraction module, configured to, when the target intent matches any reference intent, extract target parameters from the user problem and the reference data according to parameter description information associated with the any reference intent;

[0017] A sending module, configured to send the target parameters to a server through an application programming interface (API) associated with the any reference intent.

[0018] According to another aspect of the present disclosure, there is provided a user problem processing device, characterized by including:

[0019] A receiving module, configured to receive target parameters sent by a customer service system through any application programming interface (API);

[0020] A second determination module, configured to determine a target virtual machine to be processed and target operation parameters according to the target parameters and the any API;

[0021] A processing module, configured to process the target virtual machine based on the target operation parameters.

[0022] According to another aspect of the present disclosure, there is provided an electronic device, including:

[0023] At least one processor; and

[0024] A memory communicatively connected to the at least one processor; wherein,

[0025] The memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the method described in the above embodiments.

[0026] According to another aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to execute the method described in the above embodiments.

[0027] According to another aspect of the present disclosure, there is provided a computer program product, including a computer program, where when the computer program is executed by a processor, the steps of the method described in the above embodiments are implemented.

[0028] The user problem processing method, apparatus, electronic device, and storage medium provided by the present disclosure have the following beneficial effects: First, the customer service system obtains the current user problem to be processed and the reference data associated with the user problem. Then, based on the reference data, it performs intent recognition on the user problem to determine the target intent of the user problem. When the target intent matches any reference intent, it extracts the target parameters from the user problem and the reference data according to the parameter description information associated with any reference intent. Finally, it sends the target parameters to the server through the application programming interface (API) associated with any reference intent. Thus, by performing intent recognition on the reference data associated with the user problem, determining the user intent, matching the user intent with each reference intent, when the user intent matches any reference intent, determining the target parameters according to the parameter description information associated with the reference intent, and sending the target parameters to the server through the API associated with the reference intent, conditions are provided for the automated processing of user problems, effectively reducing the labor cost and improving the efficiency and accuracy of user problem processing.

[0029] It should be understood that the content described in this part is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used 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

[0030] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0031] Figure 1 is a schematic flowchart of the user problem processing method provided by an embodiment of the present disclosure;

[0032] Figure 2 is a schematic flowchart of the user problem processing method provided by another embodiment of the present disclosure;

[0033] Figure 3 is a schematic flowchart of the user problem processing method provided by another embodiment of the present disclosure;

[0034] Figure 4 is a schematic diagram of the knowledge base retrieval system architecture and the retrieval process in the user problem processing method proposed by an embodiment of the present disclosure;

[0035] Figure 5 is a schematic flowchart of the user problem processing method provided by another embodiment of the present disclosure;

[0036] Figure 6 is a schematic diagram of the architecture and process of determining the API and the target parameters in the user problem processing method proposed by an embodiment of the present disclosure;

[0037] Figure 7Schematic diagram of generating and updating Q&A pairs in the user question processing method proposed in the embodiments of the present disclosure;

[0038] Figure 8 Schematic diagram of the process of the user question processing method provided in another embodiment of the present disclosure;

[0039] Figure 9 Schematic diagram of the process of the user question processing method provided in another embodiment of the present disclosure;

[0040] Figure 10 Schematic diagram of the process of the user question processing method provided in another embodiment of the present disclosure;

[0041] Figure 11 Schematic diagram of the architecture of the user question processing method proposed in the embodiments of the present disclosure;

[0042] Figure 12 Interaction diagram of the user question processing method proposed in the embodiments of the present disclosure;

[0043] Figure 13 Schematic diagram of the structure of the user question processing device provided in an embodiment of the present disclosure;

[0044] Figure 14 Schematic diagram of the structure of the user question processing device provided in an embodiment of the present disclosure;

[0045] Figure 15 It is a block diagram of an electronic device for implementing the user question processing method of the embodiments of the present disclosure. Detailed implementation manners

[0046] The following describes exemplary embodiments of the present disclosure with reference to the accompanying drawings. Various details of the embodiments of the present disclosure are included to assist understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for clarity and conciseness, descriptions of well-known functions and structures are omitted below.

[0047] The embodiments of the present disclosure relate to the fields of artificial intelligence technologies such as large models, deep learning, natural language processing, and intelligent customer service.

[0048] Artificial Intelligence, abbreviated as AI in English. It is a new technical science that studies, develops theories, methods, technologies, and application systems for simulating, extending, and expanding human intelligence.

[0049] Large models can also be called Foundation Model models. The model extracts knowledge from hundreds of millions of corpora or images, learns, and then produces large models with hundreds of millions of parameters.

[0050] Deep Learning (DL) is to learn the internal laws and representation levels of sample data. The information obtained during these learning processes is very helpful for the interpretation of data such as text, images, and sounds. The ultimate goal of deep learning is to enable machines to have the ability to analyze and learn like humans, and be able to recognize data such as text, images, and sounds.

[0051] Natural Language Processing (NLP) is an interdisciplinary field in computer science, artificial intelligence, and linguistics. It mainly studies how to enable computers to understand, process, generate, and simulate the ability of human language, so as to achieve the ability to have natural conversations with humans.

[0052] Intelligent Customer Service is an automated service system built based on artificial intelligence, natural language processing (NLP), and machine learning technologies, aiming to solve user problems, provide information, and complete service tasks by simulating human conversations. Its core goal is to improve service efficiency, reduce labor costs, and achieve instant response.

[0053] It should be noted that in the technical solution of the present disclosure, the acquisition, storage, use, processing, etc. of data all comply with the relevant regulations of national laws and regulations and do not violate public order and good customs.

[0054] The following describes the user problem processing method, device, electronic device, and storage medium of the embodiments of the present disclosure with reference to the accompanying drawings.

[0055] Figure 1 It is a schematic flowchart of the user problem processing method provided by an embodiment of the present disclosure.

[0056] As Figure 1 shown, the user problem processing method includes:

[0057] Step 101, obtain the current user problem to be processed and the reference data associated with the user problem.

[0058] Among them, the reference data associated with the user problem can be multi-modal data related to the user problem. For example, the reference data can be text data, picture data, voice data, and user behavior logs (user clicks, dwell time, etc.), instant messages, etc. related to the user problem. The present disclosure does not make any limitations in this regard.

[0059] In the present disclosure, the customer service system can obtain reference data associated with the user's question through Websocket. Here, Websocket is a protocol for full-duplex communication over a single Transmission Control Protocol (TCP) connection, which enables real-time and two-way data exchange between the client and the server. The present disclosure does not limit this.

[0060] Step 102: Based on the reference data, perform intent recognition on the user's question to determine the target intent of the user's question.

[0061] Among them, the target intent can be the user intent corresponding to the user's question.

[0062] In the present disclosure, when performing intent recognition on the user's question based on the reference data to determine the target intent of the user's question, in the case where the reference data includes image data, text data, and user behavior log data, the image data features can be extracted through a CNN network, the text data features can be extracted through a Transformer network, and the user behavior logs can be subjected to behavior sequence analysis. Then, through a cross-modal attention mechanism, the image data features, text data features, and behavior sequences are analyzed to perform intent recognition on the user's question, determine the target intent of the user's question, and generate a target intent vector. Thus, by performing intent recognition on multi-modal data, the accuracy of intent recognition is improved.

[0063] Among them, CNN is the abbreviation of Convolutional Neural Networks.

[0064] Among them, Transformer is a neural network architecture used for natural language processing and sequence-to-sequence tasks.

[0065] Step 103: In the case where the target intent matches any reference intent, extract the target parameters from the user's question and the reference data according to the parameter description information associated with any reference intent.

[0066] Among them, the reference intent can be pre-set or can be determined as needed. For example, the reference intent can include intents such as restarting, switching, and clearing the cache of the virtual machine, etc. The present disclosure does not limit this.

[0067] Among them, the specific type and structure of the virtual machine can be determined as needed. For example, the virtual machine can be a cloud phone.

[0068] Among them, the parameter description information associated with the reference intent can be the description information of the parameters required for processing the user's question. For example, the parameter description information can include the reference device identifier, the reference order number, etc., and the present disclosure does not limit this.

[0069] It should be noted that after determining the target intent of the user's question, when matching the target intent with the reference intent, it can be directly matched based on the text, or it can also be matched based on semantics, etc., and the present disclosure does not limit this.

[0070] In the present disclosure, when the target intent matches any reference intent, by extracting the target parameters from the user's question and the reference data according to the parameter description information associated with any reference intent, the manual workload can be reduced, and the processing efficiency of the user's question can be effectively improved. For example, taking the case where the target intent matches the reference intent of "restart virtual machine" as an example, since restarting the virtual machine requires determining the device identifier of the virtual machine, therefore, the parameter description information associated with the reference intent of "restart virtual machine" can include parameter description information such as the reference device identifier. At this time, according to this parameter description information, the device identifier and other target parameters of the virtual machine to be restarted can be extracted from the user's question and the reference data, and the present disclosure does not limit this.

[0071] It should be noted that when extracting the target parameters from the user's question and the reference data, techniques such as Named Entity Recognition (NER) can be used to extract target parameters such as device identifiers and order numbers, and the present disclosure does not limit this.

[0072] Step 104, send the target parameters to the server through the application programming interface (API) associated with any reference intent.

[0073] Among them, API is the abbreviation of Application Programming Interface, and it can be used to operate the virtual machine.

[0074] It should be noted that since the reference intents are different, the operations on the virtual machine are also different, so the APIs associated with different reference intents are also different, and the present disclosure does not limit this.

[0075] In the present disclosure, after extracting the target parameters, by sending the target parameters to the server through the API associated with the reference intent that matches the target intent, the server can determine the virtual machine to be processed based on the target parameters, and solve the user's question through the API associated with the reference intent, thereby providing conditions for realizing an automated closed-loop from intent recognition to problem solving, and improving the processing efficiency of the user's question.

[0076] In the embodiments of the present disclosure, the customer service system first obtains the current user problem to be processed and the reference data associated with the user problem, then based on the reference data, performs intent recognition on the user problem to determine the target intent of the user problem, and in the case where the target intent matches any reference intent, extracts target parameters from the user problem and the reference data according to the parameter description information associated with any reference intent, and finally sends the target parameters to the server through the application programming interface (API) associated with any reference intent. Thus, by performing intent recognition on the reference data associated with the user problem, determining the user intent, matching the user intent with each reference intent, in the case where the user intent matches any reference intent, determining the target parameters according to the parameter description information associated with the reference intent, and sending the target parameters to the server through the API associated with the reference intent, conditions are provided for realizing the automated processing of user problems, effectively reducing the labor cost, and improving the efficiency and accuracy of user problem processing.

[0077] Figure 2 It is a schematic flowchart of a method for processing user problems provided in another embodiment of the present disclosure.

[0078] As Figure 2 shown, the method for processing user problems includes:

[0079] Step 201, obtain the current user problem to be processed and the reference data associated with the user problem.

[0080] Step 202, based on the reference data, perform intent recognition on the user problem to determine the target intent of the user problem.

[0081] Step 203, in the case where the target intent matches any reference intent, extract target parameters from the user problem and the reference data according to the parameter description information associated with any reference intent.

[0082] Among them, for the specific implementation forms of steps 201 to 203, reference can be made to the detailed descriptions in other embodiments of the present disclosure, and no specific elaboration will be provided here.

[0083] Step 204, in the case where complete target parameters are not extracted from the user problem and the reference data, generate a clarification question according to the unextracted parameters.

[0084] Among them, the clarification question can be used to feedback to the user which parameters need to be supplemented, so that the user can supplement the unextracted parameters.

[0085] In the present disclosure, when extracting target parameters from a user question and reference data based on parameter description information associated with a reference intention that matches the user intention, if a complete set of target parameters is not extracted, clarification questions can be generated based on the unextracted parameters for the user to supplement, so as to continue to accurately and reliably process the user question. For example, when the user intention is to restart a virtual machine, the device identifier of the virtual machine to be restarted needs to be extracted from the user question and reference data. If the device identifier is not extracted, it may not be possible to continue processing the user question. At this time, the customer service system can generate a clarification question "What is the device identifier of the virtual machine to be restarted?" based on the unextracted parameter. Herein, the specific format of the clarification question can be determined as needed, and the present disclosure does not limit this.

[0086] Step 205: Display the clarification question on the customer service interaction interface.

[0087] In the present disclosure, after generating the clarification question, the clarification question can be displayed on the customer service interaction interface, enabling the user to supplement the unextracted parameters based on the clarification question.

[0088] It should be noted that when displaying the clarification question on the customer service interaction interface, it can be displayed in any form. For example, it can be displayed in text form, or it can also be displayed in voice form, etc. The present disclosure does not limit this.

[0089] Step 206: When receiving the feedback information returned by the user, determine the unextracted parameters based on the feedback information.

[0090] The feedback information may include the unextracted parameters supplemented by the user.

[0091] In the present disclosure, after displaying the clarification question on the customer service interaction interface, when receiving the feedback information returned by the user, the unextracted parameters can be determined based on the feedback information, thereby obtaining a complete set of target parameters, providing a data basis for accurately and reliably processing the user question.

[0092] Step 207: Send the complete set of target parameters to the server via an application programming interface (API) associated with any reference intention.

[0093] The specific implementation form of step 207 is similar to that of step 104, and will not be specifically described herein.

[0094] In the embodiments of the present disclosure, the customer service system first obtains the current user problem to be processed and the reference data associated with the user problem, and based on the reference data, identifies the intention of the user problem to determine the target intention of the user problem. Then, when the target intention matches any reference intention, according to the parameter description information associated with any reference intention, the target parameters are extracted from the user problem and the reference data. When the complete target parameters cannot be extracted from the user problem and the reference data, a clarification question is generated according to the unextracted parameters. Then, the clarification question is displayed on the customer service interaction interface, and when the feedback information returned by the user is received, the unextracted parameters are determined according to the feedback information. Finally, through the application programming interface (API) associated with any reference intention, the complete target parameters are sent to the server. Thus, after determining the user intention, according to the parameter description information associated with the reference intention that matches the user intention, the target parameters are extracted from the user problem and the reference data associated with the problem. When the complete target parameters cannot be extracted, a clarification question is generated based on the unextracted parameters and displayed on the customer service interaction interface. After receiving the unextracted parameters supplemented by the user, the complete target parameters are sent to the server through the API associated with the reference intention that matches the user intention, thereby improving the efficiency and reliability of processing user problems.

[0095] Figure 3 It is a schematic flowchart of a user problem processing method provided by another embodiment of the present disclosure.

[0096] As Figure 3 shown, the user problem processing method includes:

[0097] Step 301, obtain the current user problem to be processed and the reference data associated with the user problem.

[0098] Step 302, based on the reference data, identify the intention of the user problem to determine the target intention of the user problem.

[0099] Among them, for the specific implementation forms of steps 301 to 302, reference may be made to the detailed descriptions in other embodiments of the present disclosure, and details are not described herein again.

[0100] Step 303, when the target intention does not match all reference intentions, retrieve the reference question-and-answer pairs from the knowledge base based on the target intention and the user problem.

[0101] It should be noted that the knowledge base in the present disclosure can be composed of heterogeneous data, that is, the data it contains is multi-source data. For example, taking an enterprise knowledge base as an example, the knowledge base can contain heterogeneous data such as enterprise documents, product databases, and historical work orders. Since these heterogeneous data are dynamic data, by obtaining these dynamic data in real time, the dynamic update of the knowledge base can be realized, and the timeliness of the content in the knowledge base can be improved. The present disclosure does not make any limitations in this regard.

[0102] It should be noted that the heterogeneous data in the knowledge base can construct a knowledge graph according to the relationships between entities and store it in the knowledge base in the form of a knowledge graph. The present disclosure does not make any limitations in this regard.

[0103] In the present disclosure, after determining the target intent of the user's question, when the target intent does not match all the reference intents, the target intent can be determined as a new intent. In order to determine the processing method corresponding to the target intent and improve the accuracy and reliability of processing the user's question, reference Q&A pairs can be retrieved from the knowledge base based on the target intent and the user's question.

[0104] Step 304: Determine the target reply corresponding to the user's question based on the reference Q&A pairs.

[0105] In the present disclosure, after retrieving the reference Q&A pairs from the knowledge base, the target reply corresponding to the user's question can be determined based on the reference Q&A pairs by using a pre-trained language model. For example, taking the reference Q&A pair as the question and solution method of work order #×××, where the solution method is to restart the device, the generated target reply can be "According to work order #×××, similar problems can be solved by restarting the device", so as to generate a high-precision reply through multi-modal intent recognition and dynamic knowledge base retrieval, effectively reducing the response time of the artificial customer service. The present disclosure does not make any limitations in this regard.

[0106] Among them, the specific format of the target reply can be determined according to needs. The present disclosure does not make any limitations in this regard.

[0107] Step 305: Display the target reply on the customer service interaction interface.

[0108] In the present disclosure, after determining the target reply corresponding to the user's question, the target reply can be displayed on the customer service interaction interface to provide the user with a method to solve the problem and complete the processing of the user's question. Thus, through the knowledge base retrieval system of the present disclosure, the reply to the user's question is determined, improving the practicality and timeliness of the reply to the user's question and enhancing user satisfaction.

[0109] Next, in combination with Figure 4 , an example is given to illustrate the process of determining the target reply by using the knowledge base retrieval system in the user question processing method proposed in the embodiments of the present disclosure. Figure 4Schematic diagram of the knowledge base retrieval system architecture and retrieval process in the user problem processing method proposed in the embodiments of the present disclosure.

[0110] Figure 4 In the knowledge base retrieval system in the user problem processing method of the present disclosure, through a heterogeneous data fusion engine, enterprise documents, product databases, and historical work orders are processed and fused to construct a knowledge graph, establish relationships between entities, and based on the knowledge graph, the knowledge base is dynamically updated, thereby effectively ensuring the timeliness and practicality of the knowledge in the knowledge base.

[0111] As Figure 4 shown, when determining the target reply through the knowledge base retrieval system proposed in the present disclosure, first, based on the target intention and the user problem, relevant entities in the user problem are retrieved from the knowledge base through the Retrieval Augmented Generation (RAG) method to obtain reference Q&A pairs, and based on the reference Q&A pairs, a target reply is generated using a language model, and finally the target reply is displayed on the customer service interaction interface, thereby improving the reliability and efficiency of processing user problems through the RAG technology based on the knowledge graph.

[0112] In the embodiments of the present disclosure, the customer service system first obtains the current user problem to be processed and the reference data associated with the user problem, then based on the reference data, performs intention recognition on the user problem to determine the target intention of the user problem. After that, in the case where the target intention does not match all reference intentions, based on the target intention and the user problem, reference Q&A pairs are retrieved from the knowledge base, and finally based on the reference Q&A pairs, the target reply corresponding to the user problem is determined and the target reply is displayed on the customer service interaction interface. Thus, after determining the user intention based on the reference data, in the case where the user intention does not match all reference intentions, based on the user intention and the user problem, reference Q&A pairs are retrieved from the knowledge base, and based on the reference Q&A pairs, a reply to the user problem is generated and the reply is displayed on the customer service interaction interface, thereby improving the reliability and efficiency of processing user problems.

[0113] Figure 5 Schematic diagram of the process of the user problem processing method provided in another embodiment of the present disclosure.

[0114] As Figure 5 shown, the user problem processing method includes:

[0115] Step 501, obtain the user problem submitted by the user from the customer service system.

[0116] In some possible implementation forms, the work orders to be processed in the work order system can also be parsed to determine the user problems to be processed.

[0117] In the present disclosure, the user problems submitted by users can be directly obtained through the customer service system, or the work orders to be processed in the work order system can also be parsed to determine the user problems to be processed, thereby improving the flexibility of the user problem processing method.

[0118] Step 502, obtain reference data associated with the user problem.

[0119] In the present disclosure, after determining the user problem, in order to improve the accuracy of processing the user problem, it is not only necessary to obtain data such as pictures, texts, and voices associated with the user problem sent by the user, but also necessary to determine the user's behavior log data, so that the accuracy of processing the user problem can be ensured through a dual-channel verification mechanism.

[0120] In some possible implementation forms, when obtaining reference data associated with the user problem, in order to improve the reliability of processing the user problem, it is also possible to determine the user to whom the user problem belongs, obtain the user's historical behavior log and / or historical problem information, so that the customer service system can combine historical data to determine the user's intention and improve the accuracy of intention recognition.

[0121] Step 503, based on the reference data, perform intention recognition on the user problem to determine the target intention of the user problem.

[0122] Step 504, in the case where the target intention matches any reference intention, extract target parameters from the user problem and the reference data according to the parameter description information associated with any reference intention.

[0123] Step 505, send the target parameters to the server through the application programming interface API associated with any reference intention.

[0124] Among them, for the specific implementation forms of steps 503 to 505, reference may be made to the detailed descriptions in other embodiments of the present disclosure, and no specific elaboration will be given here.

[0125] Step 506, after receiving the processing result returned by the server, display the processing result on the customer service interaction interface.

[0126] In the present disclosure, after sending the target parameters to the server and after receiving the processing result returned by the server, the processing result can be displayed on the customer service interaction interface so that the user can obtain the result in a timely manner and improve user satisfaction.

[0127] In some possible implementation forms, after receiving the processing result returned by the server, it is also possible to update the status of the work order associated with the problem to be processed in the work order system. For example, when the processing result is "resolved", the status of the work order associated with the problem to be processed can be updated to the resolved status, thereby ensuring the timeliness and accuracy of the work order system.

[0128] The following combines Figure 6 , and takes the process of determining the API and target parameters in the intention-driven automated API execution engine in the user problem processing method proposed in the embodiments of the present disclosure as an example for illustration. Figure 6 It is a schematic diagram of the architecture and process for determining the API and target parameters in the user problem processing method proposed in the embodiments of the present disclosure.

[0129] Figure 6 In, the engine architecture for determining the API and target parameters includes a semantic matching layer, a parameter extraction layer, and an automated execution layer. Among them, the automated execution layer is illustrated by taking the API call module, the supplementary parameter module, and the processing result feedback module included in Figure 6 as an example.

[0130] As Figure 6 shown, after determining the user intention based on the reference data associated with the user problem, the determined user intention can be semantically matched with all reference intentions through the semantic matching layer. In the case where there is no reference intention that matches the user intention successfully, an error prompt is returned to the customer service interaction interface, showing the reason for the matching failure. In the case where there is a reference intention that matches the user intention successfully, based on the parameter description information associated with the successfully matched reference intention, the target parameters can be extracted from the user problem and the reference data through the parameter extraction layer. In the case where complete target parameters are extracted, the API associated with the reference intention can be directly called to send the target parameters to the server. In the case where complete target parameters are not extracted, a clarification question can be generated according to the unextracted parameters, and the user is requested to supplement the unextracted parameters. Then, by calling the API associated with the reference intention, the extracted parameters and the supplemented parameters are sent to the server. Finally, the processing result returned by the server is received and the processing result is displayed on the customer service interaction interface, thus realizing the automated processing of user problems, reducing manual intervention, and improving the efficiency of user problem processing.

[0131] In the embodiments of the present disclosure, the customer service system first obtains the user questions submitted by the user from the customer service system and obtains the reference data associated with the user questions. Then, based on the reference data, it performs intent recognition on the user questions to determine the target intent of the user questions. When the target intent matches any reference intent, according to the parameter description information associated with any reference intent, it extracts the target parameters from the user questions and the reference data. After that, it sends the target parameters to the server through the application programming interface (API) associated with any reference intent. Finally, after receiving the processing result returned by the server, it displays the processing result on the customer service interaction interface. Thus, after determining the user intent through the reference data associated with the user questions, based on the parameter description information associated with the reference intent that matches the user intent, it extracts the target parameters from the user questions and the reference data, and sends the target parameters to the server through the API associated with this reference intent. After receiving the processing result returned by the server, it displays the processing result on the customer service interaction interface, thereby realizing an automated closed-loop processing from intent recognition to problem solving, improving the efficiency of processing user questions, and enhancing the user experience.

[0132] Figure 7 It is a schematic flowchart of the user question processing method provided by another embodiment of the present disclosure.

[0133] As Figure 7 shown, the user question processing method includes:

[0134] Step 701, obtain the currently to-be-processed user questions and the reference data associated with the user questions.

[0135] Step 702, based on the reference data, perform intent recognition on the user questions to determine the target intent of the user questions.

[0136] Step 703, when the target intent matches any reference intent, according to the parameter description information associated with any reference intent, extract the target parameters from the user questions and the reference data.

[0137] Among them, for the specific implementation forms of steps 701 to 703, reference can be made to the detailed descriptions in other embodiments of the present disclosure, and no specific elaboration will be provided here.

[0138] Step 704, input any reference intent and the user questions into the large model to obtain the query statement output by the large model.

[0139] It should be noted that the specific type and structure of the large model can be pre-set, or can also be determined according to actual needs. For example, the large model can be a generative large model, or a retrieval-enhanced model, or a multi-modal generative model, etc. The present disclosure does not make any limitations in this regard.

[0140] Among them, the inquiry statement can be used to confirm with the user whether the user's intention (i.e., the target intention) and the target parameters are accurate. Its specific format can be determined according to needs. For example, if the determined user intention is "insufficient device memory" and the reference intention matching the user intention is "clean cache", then combined with the user's question, the generated inquiry statement can be "Do you want to perform a cache cleaning operation on device #××?" and so on. The present disclosure does not limit this.

[0141] In the present disclosure, after extracting the target parameters from the user's question and reference data, in order to further improve the accuracy of user question processing, any reference intention matching the user intention and the user's question can be input into the large model first to obtain the inquiry statement output by the large model, so that the user can finally confirm the determined user intention and the target parameters, improving the accuracy and reliability of user question processing.

[0142] Step 705, display the inquiry statement on the customer service interaction interface.

[0143] In the present disclosure, after obtaining the inquiry statement output by the large model, the inquiry statement can be displayed on the customer service interaction interface, enabling the user to confirm the determined user intention and target parameters to ensure accurate and reliable processing of the user's question and improve user satisfaction.

[0144] Step 706, in the case of receiving the confirmation instruction returned by the user, send the target parameters to the server through the application programming interface API associated with any reference intention.

[0145] In the present disclosure, after displaying the inquiry statement on the customer service interaction interface, in the case of receiving the confirmation instruction returned by the user, it can be determined that the user intention and the target parameter recognition are accurate. At this time, the target parameters can be sent to the server through the API associated with this any reference intention. Thus, after the user confirms, the target parameters are sent to the server through the API, improving the reliability and accuracy of user question processing.

[0146] Among them, for the specific implementation form of step 706, reference can be made to the detailed description in other embodiments of the present disclosure, such as the specific implementation form in step 104, and details will not be described here again.

[0147] In the embodiments of the present disclosure, the customer service system first obtains the current user problem to be processed and the reference data associated with the user problem, and based on the reference data, performs intent recognition on the user problem to determine the target intent of the user problem. Then, when the target intent matches any reference intent, according to the parameter description information associated with any reference intent, the target parameters are extracted from the user problem and the reference data. After that, any reference intent and the user problem are input into the large model to obtain the query statement output by the large model, and the query statement is displayed on the customer service interaction interface. Finally, when the confirmation instruction returned by the user is received, the target parameters are sent to the server through the application programming interface API associated with any reference intent. Thus, after determining the user intent based on the reference data associated with the user problem, the target parameters are determined based on the parameter description information associated with the reference intent that matches the user intent, and this reference intent and the user problem are input into the large model to generate a query statement and display it on the customer service interaction interface. After the user confirms, the target parameters are sent to the server through the API associated with this reference intent, thereby improving the reliability and accuracy of user problem processing and enhancing the user experience.

[0148] Figure 8 It is a schematic flowchart of the user problem processing method provided in another embodiment of the present disclosure.

[0149] As Figure 8 shown, the user problem processing method includes:

[0150] Step 801, obtain the current user problem to be processed and the reference data associated with the user problem.

[0151] Step 802, based on the reference data, perform intent recognition on the user problem to determine the target intent of the user problem.

[0152] Step 803, when the target intent matches any reference intent, according to the parameter description information associated with any reference intent, extract the target parameters from the user problem and the reference data.

[0153] Step 804, input any reference intent and the user problem into the large model to obtain the query statement output by the large model.

[0154] Step 805, display the query statement on the customer service interaction interface.

[0155] Among them, for the specific implementation forms of steps 801 to 805, reference can be made to the detailed descriptions in other embodiments of the present disclosure, and details will not be elaborated here.

[0156] Step 806, when the correction instruction returned by the user is received, based on the correction instruction, determine the target API to be called and the updated target parameters.

[0157] Among them, the correction instruction may include the user's corrected intention and target parameters, which are not limited in this disclosure.

[0158] In this disclosure, after the inquiry statement is displayed on the customer service interface, when the corrected instruction returned by the user is received, it can be determined that there is a deviation in the target intention and / or target parameters. At this time, based on the corrected intention in the corrected instruction, the target API to be called can be determined, and the target parameters updated by the user in the corrected instruction, or based on the corrected intention, the target parameters can be updated to obtain the updated target parameters. Thus, it effectively avoids the situation of manual determination of API errors and improves the accuracy of user problem handling.

[0159] Step 807, send the updated target parameters to the server through the target API.

[0160] In this disclosure, after determining the target API to be called and the updated target parameters, the updated target parameters are sent to the server through the target API, thereby improving the accuracy of user problem handling and enhancing the user experience by the user's correction of the determined user intention and / or target parameters.

[0161] In some possible implementation forms, the parameter description information and API associated with any reference intention can also be updated based on the target API and the updated target parameters, thereby improving the practicality and timeliness of the parameter description information and API associated with any reference intention and providing a data basis for improving the accuracy of intention-driven automated operations.

[0162] In the embodiment of this disclosure, the customer service system first obtains the current user problem to be processed and the reference data associated with the user problem, and based on the reference data, performs intention recognition on the user problem to determine the target intention of the user problem. Then, when the target intention matches any reference intention, according to the parameter description information associated with any reference intention, the target parameters are extracted from the user problem and the reference data. After that, any reference intention and the user problem are input into the large model to obtain the inquiry statement output by the large model, and the inquiry statement is displayed on the customer service interaction interface. Finally, when the corrected instruction returned by the user is received, based on the corrected instruction, the target API to be called and the updated target parameters are determined, and the updated target parameters are sent to the server through the target API. Thus, after determining the user intention and the target parameters, the user intention and the user problem are input into the large model to generate an inquiry statement, which is displayed on the customer service interaction interface. When the corrected instruction returned by the user is received, based on the corrected instruction, the target API to be called and the updated target parameters are determined, and the updated target parameters are sent to the server through the target API, thereby improving the accuracy and practicality of user problem handling.

[0163] Figure 9 A schematic flowchart of a user problem processing method provided by another embodiment of the present disclosure.

[0164] As Figure 9 shown, the user problem processing method includes:

[0165] Step 901, receiving target parameters sent by a customer service system through any application programming interface (API).

[0166] In the present disclosure, when the server processes a user problem, it first receives target parameters sent by the customer service system through any API, thereby providing conditions for determining the virtual machine to be processed and the operations to be performed on the virtual machine.

[0167] Step 902, determining a target virtual machine to be processed and target operation parameters according to the target parameters and any API.

[0168] Among them, the target virtual machine may be the virtual machine corresponding to the user problem.

[0169] In some possible implementation forms, the target operation parameters may be parameters involved in operating the target virtual machine, which can be determined according to the API that sends the target parameters, and may include at least one of the following: refund amount, refund time, refund path, restart time, memory cleaning time, service identifier to be cleaned, thereby providing a data basis for processing the target virtual machine. For example, when the API that sends the target parameters is for restarting the device, the determined target operation parameters may include the restart time, etc. When the API that sends the target parameters is for cleaning the cache of the device, the determined target operation parameters may include the memory cleaning time and the service identifier to be cleaned, etc. When the API that sends the target parameters is for refunding the device, the determined target operation parameters may include the refund amount, refund time, and refund path, etc., thereby improving the accuracy of operating the target virtual machine. Among them, the specific form of the service identifier to be cleaned can be determined according to needs, such as the service name and version number, etc., and the present disclosure does not limit this.

[0170] In the present disclosure, after receiving the target parameters sent by the customer service system through any API, the target virtual machine to be processed can be determined according to the target parameters, and the target operation parameters can be determined according to the operation corresponding to the API, thereby improving the efficiency and reliability of user problem processing.

[0171] In some possible implementation forms, when determining the target virtual machine to be processed and the target operation parameters according to the target parameters and any API, the target virtual machine can be first determined according to at least one of the following in the target parameters: device identifier, user identifier, order identifier, and then the target operation parameters can be determined according to any API and the order information in the target parameters. For example, when determining the target operation parameters according to any API and the order information in the target parameters, when the operation corresponding to the API is a refund operation, the target operation parameters such as the refund amount, refund time, and refund path can be determined from the order information, thereby improving the accuracy and efficiency of determining the target virtual machine and the target operation parameters.

[0172] Among them, the device identifier, user identifier, and order identifier can be used to represent the identifiers of the device, user, and order respectively, and can be in any form. For example, the device identifier can be the device serial number, the user identifier can be the user's name, the order identifier can be the order number, etc., and the present disclosure does not limit this.

[0173] Among them, the order information can be the order information of the target virtual machine, which can include configuration parameter information such as the processor and memory of the target virtual machine, as well as relevant information on billing and payment, etc., and the present disclosure does not limit this.

[0174] Step 903, process the target virtual machine based on the target operation parameters.

[0175] In the present disclosure, after determining the virtual machine to be processed and the target operation parameters, the target virtual machine is processed based on the target operation parameters, thereby effectively avoiding the situation of manual execution of API operations going wrong, reducing costs, and improving the efficiency and accuracy of user problem handling.

[0176] In the embodiments of the present disclosure, the server first receives the target parameters sent by the customer service system through any application programming interface (API), then determines the target virtual machine to be processed and the target operation parameters according to the target parameters and any API, and finally processes the target virtual machine based on the target operation parameters. Thus, the service determines the virtual machine to be processed by receiving the target parameters sent by the customer service system through the API, determines the operation parameters for the virtual machine through this API, and then processes the virtual machine to be processed based on the operation parameters, thereby realizing an automated closed loop from user intention recognition to problem solving, effectively avoiding the error situations caused by manual operations, improving the accuracy and efficiency of user problem handling, and enhancing user satisfaction.

[0177] Figure 10 It is a schematic flowchart of a user problem handling method provided by another embodiment of the present disclosure.

[0178] Such as Figure 10As shown in the figure, the user problem processing method includes:

[0179] Step 1001: Receive the target parameters sent by the customer service system through any application programming interface (API).

[0180] Step 1002: Determine the target virtual machine to be processed and the target operation parameters according to the target parameters and any API.

[0181] Step 1003: Process the target virtual machine based on the target operation parameters.

[0182] Among them, for the specific implementation forms of steps 1001 to 1003, reference can be made to the detailed descriptions in other embodiments of the present disclosure, and no specific elaboration will be provided here.

[0183] Step 1004: Send the processing result to the customer service system through any API.

[0184] In the present disclosure, after processing the target parameter virtual machine, the processing result can be sent through any of the APIs, thereby realizing the automated solution of user problems, effectively reducing costs, and improving the efficiency of processing user problems.

[0185] In the embodiments of the present disclosure, the server first receives the target parameters sent by the customer service system through any application programming interface (API), then determines the target virtual machine to be processed and the target operation parameters according to the target parameters and any API, then processes the target virtual machine based on the target operation parameters, and finally sends the processing result to the customer service system through any API. Thus, after the server determines the virtual machine to be processed and the operation parameters based on the determined API and target parameters, it processes the virtual machine based on the operation parameters and sends the processing result to the customer service system through the API, thereby realizing the automated processing of user problems and improving the efficiency of processing user problems.

[0186] Next, in combination with Figure 11 , an example is given to illustrate the architecture of the user problem processing method proposed in the embodiments of the present disclosure. Figure 11 This is a schematic diagram of the architecture of the user problem processing method proposed in the embodiments of the present disclosure.

[0187] Figure 11 In [the figure], the reference data associated with the user problem is illustrated by taking text data, picture data, and user behavior log data as examples. The architecture of the user problem processing method includes a hybrid neural network module, a cross-modal attention mechanism module, a knowledge base retrieval module, and an intent-driven API execution engine module.

[0188] It should be noted that Figure 11 the specific architecture and retrieval process of the knowledge base retrieval module in [the figure] can be asFigure 4 as shown in the architecture and process in Figure 11 The specific architecture and execution process of the intention-driven API execution engine in Figure 6 as shown in the architecture and process in, this disclosure does not make any limitations on this.

[0189] As Figure 11 shown, after obtaining reference data such as text, pictures, and behavior log data associated with the user's question, the reference data is input into the hybrid neural network layer. The picture features are extracted through CNN, the text is encoded through Transformer and the text features are extracted, and the behavior sequence analysis is performed on the behavior log data. Then, through the cross-modal attention mechanism, the picture features, text features, and the results of the behavior sequence analysis are processed to determine the user's intention.

[0190] After that, in the case where there is a reference intention that matches the user's intention, through the intention-driven API execution engine, the API to be called and the target parameters are determined, the target parameters are sent to the server through this API, and the processing result returned by the server is received and displayed on the customer service interaction interface.

[0191] In the case where there is no reference intention that matches the user's intention, through the knowledge base retrieval system, based on the user's intention and the user's question, using the retrieval enhancement generation method, a reference Q&A pair is retrieved from the knowledge base, and based on the reference Q&A pair, the target answer corresponding to the user's question is determined and displayed on the customer service interaction interface.

[0192] Thus, through the architecture of the user question processing method proposed by this disclosure, by performing intention recognition based on multi-modal reference data, the accuracy of intention recognition is improved. After determining the user's intention, in the case where there is a reference intention that matches the user's intention, through the intention-driven API execution engine, the operations to be performed on the virtual machine and the corresponding API are determined, effectively reducing manual intervention and improving the accuracy of determining API operations. In the case where there is no reference intention that matches the user's intention, through the knowledge base retrieval system, the answer corresponding to the user's question is determined, enabling the user question processing method of this disclosure to cover more scenarios and improving the practicality of the user question processing method. Therefore, through the architecture of the user question processing method proposed by this disclosure, an automated closed-loop from intention recognition to problem solving is achieved, shortening the response time for processing user questions and improving the processing efficiency and accuracy of user questions.

[0193] Next, in combination with Figure 12 , an example is given to illustrate the interaction process of the user question processing method proposed in the embodiments of this disclosure. Figure 12 is the interaction schematic diagram of the user question processing method proposed in the embodiments of this disclosure.

[0194] Step 1201, the customer service system obtains the current user problem to be processed and the reference data associated with the user problem.

[0195] Step 1202, the customer service system performs intent recognition on the user problem based on the reference data to determine the target intent of the user problem.

[0196] Step 1203, when the target intent matches any reference intent, the customer service system extracts target parameters from the user problem and the reference data according to the parameter description information associated with the any reference intent.

[0197] Step 1204, the customer service system sends the target parameters to the server through the application programming interface API associated with the any reference intent.

[0198] In some possible implementation forms, the server receives the target parameters sent by the customer service system through the any application programming interface API.

[0199] Step 1205, the server determines the target virtual machine to be processed and the target operation parameters according to the target parameters and the any API.

[0200] Step 1206, the server processes the target virtual machine based on the target operation parameters.

[0201] Step 1207, the server sends the processing result to the customer service system through the any API.

[0202] In some possible implementation forms, the customer service system receives the processing result sent by the server through the any API.

[0203] Step 1208, the customer service system displays the processing result on the customer service interaction interface.

[0204] Among them, for the specific implementation forms of steps 1201 to 1208, reference can be made to the detailed descriptions in other embodiments of the present disclosure, and details are not described herein again.

[0205] In the embodiments of the present disclosure, through the user problem processing method of the present disclosure, the automated interaction between the customer service system and the server is realized, effectively shortening the user problem processing time and improving the efficiency of user problem processing.

[0206] To implement the above embodiments, the embodiments of the present disclosure also propose a user problem processing device.

[0207] Figure 13 It is a schematic structural diagram of a user problem processing device provided by an embodiment of the present disclosure.

[0208] As Figure 13As shown, the user problem processing device 1300 includes: an acquisition module 1301, a first determination module 1302, an extraction module 1303, and a sending module 1304.

[0209] The acquisition module 1301 is used to acquire the user problem to be processed currently and the reference data associated with the user problem.

[0210] The first determination module 1302 is used to perform intent recognition on the user problem based on the reference data to determine the target intent of the user problem.

[0211] The extraction module 1303 is used to extract target parameters from the user problem and the reference data according to the parameter description information associated with any reference intent when the target intent matches any reference intent.

[0212] The sending module 1304 is used to send the target parameters to the server through the application programming interface API associated with any reference intent.

[0213] Optionally, the above-mentioned extraction module 1303 is specifically used for:

[0214] When the complete target parameters cannot be extracted from the user problem and the reference data, generate a clarification question according to the unextracted parameters.

[0215] Display the clarification question on the customer service interaction interface.

[0216] When receiving the feedback information returned by the user, determine the unextracted parameters according to the feedback information.

[0217] Optionally, the above-mentioned first determination module 1302 is further used for:

[0218] When the target intent does not match all reference intents, retrieve the reference Q&A pairs from the knowledge base based on the target intent and the user problem.

[0219] Based on the reference Q&A pairs, determine the target reply corresponding to the user problem.

[0220] Display the target reply on the customer service interaction interface.

[0221] Optionally, the above-mentioned acquisition module 1301 is further used for any one of the following:

[0222] Acquire the user problem submitted by the user from the customer service system.

[0223] Parse the work order to be processed in the work order system to determine the user problem to be processed.

[0224] Optionally, the above-mentioned sending module 1304 is further used for:

[0225] After receiving the processing result returned by the server, display the processing result on the customer service interaction interface.

[0226] Optionally, the above-mentioned sending module 1304 is further configured to:

[0227] After receiving the processing result returned by the server, update the status of the work order associated with the problem to be processed in the work order system.

[0228] Optionally, the above-mentioned obtaining module 1301 is specifically configured to:

[0229] Determine the user to whom the user problem belongs;

[0230] Obtain the historical behavior log and / or historical problem information of the user.

[0231] Optionally, the above-mentioned sending module 1304 is specifically configured to:

[0232] When the target intention matches any reference intention, input any reference intention and the user problem into the large model to obtain the query statement output by the large model;

[0233] Display the query statement on the customer service interaction interface;

[0234] When receiving the confirmation instruction returned by the user, send the target parameter to the server.

[0235] Optionally, the above-mentioned sending module 1304 is further configured to:

[0236] When receiving the correction instruction returned by the user, based on the correction instruction, determine the target API to be called and the updated target parameter;

[0237] Send the updated target parameter to the server through the target API.

[0238] Optionally, the above-mentioned sending module 1304 is further configured to:

[0239] Based on the target API and the updated target parameter, update the parameter description information and API associated with any reference intention.

[0240] It should be noted that the explanations of the foregoing embodiments of the user problem processing method also apply to the user problem processing device of this embodiment, so they will not be repeated here.

[0241] In an embodiment of the present disclosure, the customer service system first obtains the current user problem to be processed and the reference data associated with the user problem, then based on the reference data, performs intent recognition on the user problem to determine the target intent of the user problem, and in the case where the target intent matches any reference intent, extracts target parameters from the user problem and the reference data according to the parameter description information associated with any reference intent, and finally sends the target parameters to the server through the application programming interface (API) associated with any reference intent. Thus, by performing intent recognition on the reference data associated with the user problem, determining the user intent, matching the user intent with each reference intent, in the case where the user intent matches any reference intent, determining the target parameters according to the parameter description information associated with the reference intent, and sending the target parameters to the server through the API associated with the reference intent, conditions are provided for realizing the automated processing of user problems, effectively reducing the labor cost and improving the efficiency and accuracy of processing user problems.

[0242] Figure 14 FIG. is a schematic structural diagram of a user problem processing device provided in an embodiment of the present disclosure.

[0243] As Figure 14 shown, the user problem processing device 1400 includes: a receiving module 1401, a second determination module 1402, and a processing module 1403.

[0244] The receiving module 1401 is configured to receive the target parameters sent by the customer service system through any application programming interface (API);

[0245] The second determination module 1402 is configured to determine the target virtual machine to be processed and the target operation parameters according to the target parameters and any API;

[0246] The processing module 1403 is configured to process the target virtual machine based on the target operation parameters.

[0247] Optionally, the second determination module 1402 is specifically configured to:

[0248] Determine the target virtual machine according to at least one of the following in the target parameters: device identifier, user identifier, order identifier;

[0249] Determine the target operation parameters according to any API and the order information in the target parameters.

[0250] Optionally, the target operation parameters include at least one of the following: refund amount, refund time, refund path, restart time, memory cleaning time, service identifier to be cleaned.

[0251] Optionally, the processing module 1403 is further configured to:

[0252] Send the processing result to the customer service system through any API.

[0253] It should be noted that the explanations of the embodiments of the foregoing user problem processing method are also applicable to the user problem processing device of this embodiment, so they will not be repeated here.

[0254] In the embodiments of the present disclosure, the server first receives target parameters sent by the customer service system through any application programming interface (API), then determines the target virtual machine to be processed and the target operation parameters according to the target parameters and any API, and finally processes the target virtual machine based on the target operation parameters. Thus, the service determines the virtual machine to be processed by receiving the target parameters sent by the customer service system through the API, determines the operation parameters for the virtual machine through this API, and then processes the virtual machine to be processed based on the operation parameters, thereby realizing an automated closed loop from user intention recognition to problem solving, effectively avoiding error situations caused by manual operations, improving the accuracy and efficiency of user problem processing, and enhancing user satisfaction.

[0255] According to the embodiments of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0256] Figure 15 FIG. shows a schematic block diagram of an exemplary electronic device 1500 that can be used to implement the embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely exemplary and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0257] As Figure 15 shown, the device 1500 includes a computing unit 1501, which can execute various appropriate actions and processes according to the computer program stored in a ROM (Read-Only Memory) 1502 or the computer program loaded from a storage unit 1508 into a RAM (Random Access Memory) 1503. In the RAM 1503, various programs and data required for the operation of the device 1500 can also be stored. The computing unit 1501, the ROM 1502, and the RAM 1503 are connected to each other through a bus 1504. An I / O (Input / Output) interface 1505 is also connected to the bus 1504.

[0258] A plurality of components in device 1500 are connected to I / O interface 1505, including: an input unit 1506, such as a keyboard, a mouse, etc.; an output unit 1507, such as various types of displays, speakers, etc.; a storage unit 1508, such as a disk, an optical disc, etc.; and a communication unit 1509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1509 allows device 1500 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0259] The computing unit 1501 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 1501 include but are not limited to a CPU (Central Processing Unit), a GPU (Graphic Processing Units), various dedicated AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, a DSP (Digital Signal Processor), and any suitable processor, controller, microcontroller, etc. The computing unit 1501 executes the various methods and processes described above, such as the user problem processing method. For example, in some embodiments, the user problem processing method can be implemented as a computer software program that is tangibly contained in a machine-readable medium, such as the storage unit 1508. In some embodiments, part or all of the computer program can be loaded and / or installed onto device 1500 via the ROM 1502 and / or the communication unit 1509. When the computer program is loaded into the RAM 1503 and executed by the computing unit 1501, one or more steps of the user problem processing method described above can be executed. Alternatively, in other embodiments, the computing unit 1501 can be configured to execute the user problem processing method in any other suitable manner (e.g., by means of firmware).

[0260] The various embodiments of the systems and techniques described above in this document can be implemented in digital electronic circuitry, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application Specific Standard Products), SoCs (System On Chip), CPLDs (Complex Programmable Logic Devices), 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 the data and instructions to the storage system, the at least one input device, and the at least one output device.

[0261] The program code for implementing the methods of the present disclosure can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the program codes are executed by the processor or controller, the functions / operations specified in the flowchart and / or block diagram are implemented. The program code can be executed entirely on the machine, partially on the machine, as an independent software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0262] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. The machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a RAM, a ROM, an EPROM (Electrically Programmable Read-Only-Memory), or a flash memory, an optical fiber, a CD-ROM (Compact Disc Read-Only Memory), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0263] In order to provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or an LCD (Liquid Crystal Display) monitor); and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the computer. Other kinds 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, speech input, or tactile input).

[0264] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected to each other by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a LAN (Local Area Network), a WAN (Wide Area Network), the Internet, and a blockchain network.

[0265] A computer system may include a client and a server. The client and the server are generally far from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, and solves the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services (Virtual Private Server). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0266] According to an embodiment of the present disclosure, the present disclosure also provides a computer program product, which, when executed by an instruction processor in the computer program product, executes the user problem processing method proposed in the above embodiments of the present disclosure.

[0267] It should be understood that various forms of the processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, and no limitations are imposed herein.

[0268] The above specific embodiments do not constitute a limitation on the protection scope of the present disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present disclosure shall be included within the protection scope of the present disclosure.

Claims

1. A method for processing user problems, characterized in that, Including: Obtain the current user question to be processed and the reference data associated with the user question; Based on the reference data, perform intent recognition on the user question to determine the target intent of the user question; In the case where the target intent matches any reference intent, extract target parameters from the user question and the reference data according to the parameter description information associated with the any reference intent; Send the target parameters to the server through the application programming interface API associated with the any reference intent.

2. The method according to claim 1, characterized in that The extracting target parameters from the user question and the reference data according to the parameter description information associated with the any reference intent includes: In the case where complete target parameters are not extracted from the user question and the reference data, generate a clarification question according to the unextracted parameters; Display the clarification question on the customer service interaction interface; In the case where feedback information returned by the user is received, determine the unextracted parameters according to the feedback information.

3. The method according to claim 1, characterized in that After determining the target intent of the user question, it further includes: In the case where the target intent does not match all reference intents, retrieve reference question-and-answer pairs from the knowledge base based on the target intent and the user question; Based on the reference question-and-answer pairs, determine the target reply corresponding to the user question; Display the target reply on the customer service interaction interface.

4. The method according to claim 1, wherein The obtaining the current user question to be processed includes any one of the following: Obtain the user question submitted by the user from the customer service system; Parse the work order to be processed in the work order system to determine the user question to be processed.

5. The method according to claim 4, wherein After sending the target parameters to the server, it further includes: After receiving the processing result returned by the server, display the processing result on the customer service interaction interface.

6. The method according to claim 4, wherein After sending the target parameters to the server, it further includes: After receiving the processing result returned by the server, update the status of the work order associated with the question to be processed in the work order system.

7. The method according to claim 4, wherein Obtain the reference data associated with the user question, including: Determine the user to whom the user question belongs; Obtain the historical behavior log and / or historical question information of the user.

8. The method according to any one of claims 1-7, characterized in that, The sending the target parameters to the server includes: In the case where the target intent matches any reference intent, input the any reference intent and the user question into the large model to obtain the query statement output by the large model; Display the query statement on the customer service interaction interface; In the case where a confirmation instruction returned by the user is received, send the target parameters to the server.

9. The method according to claim 8, wherein After displaying the query statement on the customer service interaction interface, it further includes: In the case where a correction instruction returned by the user is received, determine the target API to be called and the updated target parameters based on the correction instruction; Send the updated target parameters to the server through the target API.

10. The method according to claim 9, wherein After determining the target API to be called and the updated target parameters based on the correction instruction, it further includes: Update the parameter description information and API associated with any reference intent based on the target API and the updated target parameters.

11. A method for processing user problems, characterized in that, Including: Receive target parameters sent by the customer service system through any application programming interface (API). Determine the target virtual machine to be processed and the target operation parameters according to the target parameters and the any API. Process the target virtual machine based on the target operation parameters.

12. The method according to claim 11, wherein The determining the target virtual machine to be processed and the target operation parameters according to the target parameters and the any API includes: Determine the target virtual machine according to at least one of the following in the target parameters: device identifier, user identifier, order identifier. Determine the target operation parameters according to the any API and the order information in the target parameters.

13. The method according to claim 11, characterized in that, The target operation parameters include at least one of the following: refund amount, refund time, refund path, restart time, memory cleaning time, service identifier to be cleaned.

14. The method according to any one of claims 11-13, characterized in that, After processing the target virtual machine, it further includes: Send the processing result to the customer service system through the any API.

15. A user problem processing device, characterized in that, Including: An acquisition module, configured to acquire the current user problem to be processed and the reference data associated with the user problem. A first determination module, configured to perform intent recognition on the user problem based on the reference data and determine the target intent of the user problem. An extraction module, configured to extract target parameters from the user problem and the reference data according to the parameter description information associated with any reference intent when the target intent matches any reference intent. A sending module, configured to send the target parameters to the server through the application programming interface (API) associated with any reference intent.

16. A user problem processing device, characterized in that, Including: A receiving module, configured to receive target parameters sent by the customer service system through any application programming interface (API). A second determination module, configured to determine the target virtual machine to be processed and the target operation parameters according to the target parameters and the any API. A processing module, configured to process the target virtual machine based on the target operation parameters.

17. An electronic device, including: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the method according to any one of claims 1-14.

18. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are used to cause the computer to execute the method according to any one of claims 1-14.

19. A computer program product, including a computer program, where the computer program implements the steps of the method according to any one of claims 1-14 when executed by a processor.

Images