Consultation processing method and device, storage medium and electronic equipment

Abstract

The invention discloses a consultation processing method and device, a storage medium and electronic equipment. The method comprises the steps that a consultation request of a user is received, and the consultation request is a consultation request associated with system operation and maintenance; inputting the consultation request into a target model to obtain a consultation reply, the target model being obtained by training a plurality of groups of training samples, and each group of training samples comprising a historical consultation request and a historical consultation reply; determining a target field to which the consultation request belongs, and calculating the similarity between the consultation reply and the data of the expert knowledge base of the target field; and when the similarity is greater than or equal to a similarity threshold, feeding back the consultation reply to the client of the user. Through the method and the device, the problem of low consultation processing efficiency in the field of system operation and maintenance in related technologies is solved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and more particularly, to a consultation processing method, apparatus, storage medium, and electronic device. Background Art

[0002] As the business of the business system increases day by day, the consultation support related to system operation and maintenance that needs to be provided also increases accordingly. The system operation and maintenance information is highly specialized, and the user's demand for accurate and professional consultation support is increasing day by day.

[0003] In the related art, in the consultation processing method for the system operation and maintenance field, one is to adopt manual consultation support, but the labor cost and time cost of manual consultation support are relatively large; the other is to provide pre-edited consultation documents to solve the user's consultation, but this method lacks interactivity, and some consultation problems still cannot be solved.

[0004] In view of the problem of low consultation processing efficiency in the related art for the system operation and maintenance field, no effective solution has been proposed yet. Summary of the Invention

[0005] The main object of the present application is to provide a consultation processing method, apparatus, storage medium, and electronic device to solve the problem of low consultation processing efficiency in the related art for the system operation and maintenance field.

[0006] To achieve the above object, according to one aspect of the present application, a consultation processing method is provided. The method includes: receiving a consultation request from a user, where the consultation request is a consultation request associated with system operation and maintenance; inputting the consultation request into a target model to obtain a consultation response, where the target model is trained by multiple sets of training samples, and each set of training samples includes a historical consultation request and a historical consultation response; determining a target field to which the consultation request belongs, and calculating the similarity between the consultation response and the data in the expert knowledge base of the target field; and in the case where the similarity is greater than or equal to a similarity threshold, feeding back the consultation response to the client of the user.

[0007] Optionally, determining the target field to which the consultation request belongs includes: extracting a set of keywords from the consultation request, inputting the set of keywords into a clustering model to obtain a clustering result; determining the probability that the consultation request belongs to each field from the clustering result to obtain a set of probabilities; and determining the field corresponding to the maximum probability in the set of probabilities as the target field.

[0008] Optionally, calculating the similarity between the consultation reply and the data in the expert knowledge base of the target field includes: preprocessing the consultation reply to obtain a preprocessed reply statement, where the preprocessing includes at least one of the following: removing stop words, converting punctuation and case, and lemmatization; vectorizing the reply statement to obtain a target word vector; extracting multiple document vectors from the expert knowledge base and calculating the cosine similarity between the target word vector and each document vector to obtain the similarity between the consultation reply and the data in the expert knowledge base of the target field.

[0009] Optionally, after calculating the similarity between the consultation reply and the data in the expert knowledge base of the target field, the method further includes: in the case where the similarity is less than the similarity threshold, obtaining the background description of the target field; adding the background description to the consultation request to obtain an updated consultation request, inputting the updated consultation request into the target model to obtain an updated consultation reply; in the case where the similarity between the updated consultation reply and the data in the expert knowledge base is greater than or equal to the similarity threshold, feeding back the updated consultation reply to the user's client; in the case where the similarity between the updated consultation reply and the data in the expert knowledge base is less than the similarity threshold, sending a prompt message, where the prompt message is used to prompt the user to increase the description content of the updated consultation request.

[0010] Optionally, the target model is trained in the following manner: obtaining the expert knowledge data in the system operation and maintenance field and preprocessing the expert knowledge data to obtain preprocessed expert knowledge data, where the preprocessing includes at least one of the following: deduplication, noise reduction, text cleaning, and tokenization and encoding; converting the preprocessed expert knowledge data into a word vector matrix through a decoder and adding position information encoding to the word vector matrix to obtain an updated word vector matrix; determining an initial natural language processing model based on the updated word vector matrix; obtaining the historical consultation requests and historical consultation replies in the system operation and maintenance field and determining each historical consultation request and the corresponding historical consultation reply as a set of training samples to obtain multiple sets of training samples; training the initial natural language processing model based on the multiple sets of training samples to obtain the target model.

[0011] Optionally, the method further includes: updating the expert knowledge data in the system operation and maintenance field at preset intervals to obtain updated expert knowledge data; updating the word vector matrix of the target model based on the updated expert knowledge data to obtain an updated target model.

[0012] Optionally, before determining the target field to which the consultation request belongs, the method further includes: determining multiple systems and databases applied by the target institution, determining a field based on the operation and maintenance data of each system or database to obtain multiple fields; for each field, collecting the professional terms, operation and maintenance problems, and problem replies of the field; constructing an expert knowledge base of the field based on the professional terms, operation and maintenance problems, and problem replies.

[0013] To achieve the above object, according to another aspect of the present application, a consultation processing device is provided. The device includes: a receiving unit, configured to receive a consultation request from a user, where the consultation request is a consultation request associated with system operation and maintenance; an input unit, configured to input the consultation request into a target model to obtain a consultation reply, where the target model is trained by multiple sets of training samples, and each set of training samples includes a historical consultation request and a historical consultation reply; a calculation unit, configured to determine a target field to which the consultation request belongs, and calculate the similarity between the consultation reply and the data in the expert knowledge base of the target field; a first feedback unit, configured to, when the similarity is greater than or equal to a similarity threshold, feedback the consultation reply to the client of the user.

[0014] In an embodiment of the present application, the method includes receiving a consultation request from a user, where the consultation request is a consultation request associated with system operation and maintenance; inputting the consultation request into a target model to obtain a consultation reply, where the target model is trained by multiple sets of training samples, and each set of training samples includes a historical consultation request and a historical consultation reply; determining a target field to which the consultation request belongs, and calculating the similarity between the consultation reply and the data in the expert knowledge base of the target field; and when the similarity is greater than or equal to a similarity threshold, feedbacking the consultation reply to the client of the user. By training the target model, generating a consultation reply to the consultation request by the target model, and when the similarity between the consultation reply and the data in the expert knowledge base of the target field is greater than or equal to the similarity threshold, feedbacking the consultation reply to the user, the purpose of automatically solving daily consultations in the field of system operation and maintenance is achieved, thereby achieving the technical effect of improving the processing efficiency of consultations, and further solving the technical problem of low processing efficiency of consultations in the field of system operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The accompanying drawings, which form a part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions thereof of this application are used to explain this application and do not constitute an improper limitation to this application. In the drawings:

[0016] Figure 1 shows a hardware structure block diagram of a computer terminal (or mobile device) for implementing the consultation processing method;

[0017] Figure 2 is a flowchart of the consultation processing method provided by an embodiment of the present application;

[0018] Figure 3 is a schematic diagram of the consultation processing device provided by an embodiment of the present application;

[0019] Figure 4 is a structure block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] In order to enable those skilled in the art to better understand the solution of this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data used can be interchanged under appropriate circumstances, so that the embodiments of this application described here can be implemented in an order different from those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices.

[0022] It should be noted that the information collected in this application (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) are information and data authorized by the user or fully authorized by all parties. And the processing of relevant data, such as collection, storage, use, processing, transmission, provision, disclosure, and application, all comply with relevant laws, regulations, and standards, take necessary confidentiality measures, do not violate public order and good customs, and provide corresponding operation entrances for users to choose to authorize or refuse. For example, an interface is set between this system and relevant users or institutions to provide corresponding operation entrances for users to choose to agree or refuse the results of automated decision-making; if the user chooses to refuse, the expert decision-making process will be entered.

[0023] Embodiment 1

[0024] According to the embodiments of this application, an embodiment of a method for consulting processing is also provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described here can be executed in an order different from that here.

[0025] The method embodiment provided by the first embodiment of this application can be executed on a mobile terminal, a computer terminal, or a similar computing device. Figure 1The figure shows a hardware block diagram of a computer terminal (or mobile device) for implementing a consultation processing method. As Figure 1 shown, the computer terminal 10 (or mobile device) may include one or more processors 102 (illustrated as 102a, 102b, ……, 102n in the figure) (the processor 102 may include, but is not limited to, processing devices such as an MCU (Microcontroller Unit, microprocessor) or an FPGA (Field-Programmable Gate Array, programmable logic device)), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may further include: a display, an input / output interface (I / O interface), a USB (Universal Serial Bus, universal serial bus) port (which may be included as one of the ports of the BUS (Business, bus)), a network interface, a power supply, and / or a camera. Those of ordinary skill in the art can understand that Figure 1 the structure shown is only schematic and does not limit the structure of the above-mentioned electronic device. For example, the computer terminal 10 may further include more or fewer components than Figure 1 shown, or have a different configuration from Figure 1 shown.

[0026] It should be noted that the above one or more processors 102 and / or other data processing circuits are generally referred to as "data processing circuits" herein. The data processing circuit may be embodied in software, hardware, firmware, or any combination thereof, in whole or in part. In addition, the data processing circuit may be a single independent processing module, or be incorporated in whole or in part into any one of the other elements in the computer terminal 10 (or mobile device). As involved in the embodiments of the present application, the data processing circuit is a kind of processor control (such as the selection of a variable resistance terminal path connected to an interface).

[0027] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the consultation processing method in the embodiments of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, implements the above-mentioned consultation processing method. The memory 104 may include a high-speed random access memory, and may further include a non-volatile memory, such as one or more magnetic storage devices, a flash memory, or other non-volatile solid-state memories. In some instances, the memory 104 may further include a memory remotely set relative to the processor 102, and these remote memories can be connected to the computer terminal 10 through a network. Examples of the above network include, but are not limited to, the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0028] The transmission device 106 is used to receive or send data via a network. Specific examples of the above-mentioned network may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a Radio Frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0029] The display can be, for example, a touch-screen liquid crystal display, which enables the user to interact with the user interface of the computer terminal 10 (or mobile device).

[0030] Under the above operating environment, the present application provides a consultation processing method. Figure 2 It is a flowchart of the consultation processing method provided by the embodiment of the present application. As Figure 2 shown, the method includes:

[0031] Step S201, receiving a consultation request from a user, where the consultation request is a consultation request associated with system operation and maintenance.

[0032] In step S201, in the consultation processing system applying the consultation processing method provided by this embodiment, a consultation support function is provided on the system home page. The front end can use a front-end framework for building a user interface to implement the support consultation page effect, and the back end can be implemented through a container component framework. The user interface of the system can be a web page, a mobile application or an application programming interface, and is used to receive the consultation request from the user. The user asks questions through input or voice, and the system needs to be able to accurately receive and parse these input consultation requests.

[0033] After receiving the consultation request from the user, the system performs preprocessing, which can include text cleaning (removing special characters, punctuation marks), word segmentation, stemming, etc., to convert the text into a machine-processable format. Then, natural language understanding technology is used to parse the user's intention, identify the key information and field of the consultation request, and judge whether the request is related to system operation and maintenance. If it is related, continue to the next step of processing; if it is not related, prompt the user that the input consultation request is incorrect and needs to be re-entered.

[0034] Step S202, inputting the consultation request into a target model to obtain a consultation reply, where the target model is obtained by training with multiple groups of training samples, and each group of training samples includes a historical consultation request and a historical consultation reply.

[0035] In step S202, the target model can be a natural language processing model pre-trained with expert knowledge in the field of system operation and maintenance. By collecting a large number of consultation requests related to system operation and maintenance and their corresponding consultation responses as training samples for model training. Based on the Transformer architecture (a neural network architecture based on the attention mechanism, that is, a natural language processing model), the training samples are used for training. The training objective is to learn the mapping relationship that converts consultation requests into consultation responses. The target model captures the dependencies between words in long sequences through the multi-head attention mechanism, enabling the model to understand the context and details of the consultation requests.

[0036] After receiving a new consultation request, it is preprocessed and converted into a word vector representation. The preprocessed request is input into the trained target model, and the target model uses the knowledge and patterns it has learned to generate a consultation response.

[0037] Step S203, determine the target field to which the consultation request belongs, and calculate the similarity between the consultation response and the data in the expert knowledge base of the target field.

[0038] In step S203, first, use natural language processing technology to analyze the consultation request to identify and determine the target field to which it belongs. For example, this can be achieved by training a clustering model for field classification. The clustering model can classify the consultation request into the correct target field based on the keywords in the consultation request. The classified fields can be pre-annotated. For example, in a financial institution, various business systems and databases that need to be maintained, and the operation and maintenance issues related to each business system or database are regarded as a field, and an expert knowledge base for each field is constructed. The expert knowledge base can contain information such as professional documents, common questions and answers, and standard solutions in the target field. This information is vectorized using the same vectorization method as the consultation request to facilitate comparison in the vector space.

[0039] After determining the target field to which the consultation request belongs, convert the consultation request into a numerical vector. After determining the vector representations of the consultation request and the data in the expert knowledge base, the similarity between the two can be calculated. According to expert experience, a similarity threshold is set. Only when the similarity is higher than this similarity threshold, it indicates that the accuracy of the consultation response output by the target model is relatively high, and at this time, the consultation response will be fed back to the user.

[0040] Step S204, in the case where the similarity is greater than or equal to the similarity threshold, feedback the consultation response to the user's client.

[0041] In step S204, the similarity threshold can be set based on the expected accuracy and relevance of the consultation response. If the calculated similarity reaches or exceeds this threshold, it indicates that the consultation response is sufficiently relevant and accurate and can be provided to the user. Convert the selected consultation response into a text format readable by the user. The system sends the consultation response to the user's client through the user interface.

[0042] It should be noted that after receiving the consultation response, the user can further ask questions or give feedback. Through the user feedback mechanism, the user is allowed to evaluate the quality of the consultation response, which helps the system learn and optimize, and improve the accuracy of future consultation processing. The system can record each consultation request and response, as well as the user's feedback, for subsequent model training and update of the expert knowledge base. By collecting and analyzing this data, the similarity threshold can be continuously adjusted, the model parameters of the target model can be optimized, and the structure of the expert knowledge base can be improved, thereby improving the overall performance of the system and the accuracy of the consultation response.

[0043] The consultation processing method provided by the embodiments of the present application receives a consultation request from a user, where the consultation request is a consultation request associated with system operation and maintenance; inputs the consultation request into a target model to obtain a consultation response, where the target model is trained by multiple sets of training samples, and each set of training samples includes a historical consultation request and a historical consultation response; determines the target field to which the consultation request belongs, and calculates the similarity between the consultation response and the data in the expert knowledge base of the target field; in the case where the similarity is greater than or equal to the similarity threshold, feedback the consultation response to the user's client. By training the target model, generating the consultation response of the consultation request by the target model, and when the similarity between the consultation response and the data in the expert knowledge base of the target field is greater than or equal to the similarity threshold, feedback the consultation response to the user, it achieves the purpose of automatically solving the daily consultations in the field of system operation and maintenance, thereby realizing the technical effect of improving the processing efficiency of consultations, and further solving the technical problem of low processing efficiency of consultations in the field of system operation and maintenance.

[0044] To evaluate the accuracy of the consultation response, it is necessary to first determine the target field described in the consultation request. Optionally, in the consultation processing method provided by the embodiments of the present application, determining the target field to which the consultation request belongs includes: extracting a set of keywords from the consultation request, inputting the set of keywords into a clustering model to obtain a clustering result; determining the probability that the consultation request belongs to each field from the clustering result to obtain a set of probabilities; and determining the field corresponding to the maximum probability in the set of probabilities as the target field.

[0045] In some examples, first, text preprocessing is performed on the consultation request. The preprocessing may include word segmentation, stop word removal, stemming, etc., to reduce noise and focus on key information. Then, a keyword extraction algorithm is used to screen out a set of keywords from the preprocessed text. These keywords are used to characterize the core content and domain features of the consultation request.

[0046] The set of extracted keywords is input into a pre-trained clustering model. The clustering model can be trained based on unsupervised learning algorithms such as K-means clustering, hierarchical clustering, etc. The clustering model will calculate the correlation or similarity between the consultation request and each predefined domain based on the vector representation of the set of keywords. Determine the probability that the consultation request belongs to each domain. From the calculated set of probabilities, the domain corresponding to the maximum probability is selected as the target domain of the consultation request.

[0047] In this embodiment, through the combination of keyword extraction and the clustering model, the system can automatically infer the domain attribution of the consultation request without explicit domain annotation, and then evaluate the accuracy of the consultation reply based on the data in the expert knowledge base of the target domain, improving the processing efficiency and accuracy of the consultation reply.

[0048] The similarity can be calculated based on cosine similarity. Optionally, in the consultation processing method provided in the embodiments of the present application, calculating the similarity between the consultation reply and the data in the expert knowledge base of the target domain includes: preprocessing the consultation reply to obtain a preprocessed reply statement, where the preprocessing includes at least one of the following: removing stop words, punctuation, case conversion, and lemmatization; vectorizing the reply statement to obtain a target word vector; extracting multiple document vectors from the expert knowledge base and calculating the cosine similarity between the target word vector and each document vector to obtain the similarity between the consultation reply and the data in the expert knowledge base of the target domain.

[0049] In some examples, the stop words can be high-frequency but meaningless words such as "de", "shi", and "zai" in the consultation reply to reduce noise and highlight key information. By unifying or removing punctuation in the consultation reply, the consistency and conciseness of the text are maintained. Lemmatization refers to processing both "running" and "ran" as "run" to enhance semantic consistency.

[0050] The response statement is vectorized through a word embedding model to obtain a target word vector, and the documents in the expert knowledge base can be obtained as document vectors through the same processing method. The cosine similarity between each document vector and the target word vector is calculated as the similarity between the consultation response and the data in the expert knowledge base of the target domain. When the cosine similarity between at least one document vector and the target word vector is greater than or equal to the similarity threshold, it can be determined that the similarity between the consultation response and the data in the expert knowledge base of the target domain meets the conditions.

[0051] In this embodiment, by calculating the similarity between the consultation response and the data in the expert knowledge base of the target domain, it is ensured that the consultation response output by the target model conforms to the target domain, thereby improving the quality and accuracy of the consultation response.

[0052] To improve the accuracy of the consultation response, a background description of the target domain can be added to the consultation request. Optionally, in the consultation processing method provided in the embodiments of the present application, after calculating the similarity between the consultation response and the data in the expert knowledge base of the target domain, the method further includes: when the similarity is less than the similarity threshold, obtaining the background description of the target domain; adding the background description to the consultation request to obtain an updated consultation request, inputting the updated consultation request into the target model to obtain an updated consultation response; when the similarity between the updated consultation response and the data in the expert knowledge base is greater than or equal to the similarity threshold, feeding back the updated consultation response to the user's client; when the similarity between the updated consultation response and the data in the expert knowledge base is less than the similarity threshold, sending a prompt message, where the prompt message is used to prompt the user to increase the description content of the updated consultation request.

[0053] In some examples, if the similarity is less than the similarity threshold, it means that the correlation between the consultation response and the target domain is low, that is, the accuracy of the consultation response is poor. The correlation between the consultation response output by the target model and the target domain can be improved by supplementing the background description. The background description can be the general definition of the target domain, an overview of common problems, an explanation of technical terms, etc., aiming to provide richer context information for the consultation request. Combining the obtained background description with the original consultation request forms an updated consultation request. The combination method can be directly attaching the background description after the original request, or integrating the key information of the background description into the request to make it more specific and rich.

[0054] The updated consultation request text is input into the target model for deep learning inference to generate an updated consultation response. Since the target model receives more comprehensive information of the updated consultation request, it can provide an updated consultation response based on a broader context, improving the accuracy and relevance of the response. In the same way as the processing of the original consultation request, the system calculates the similarity between the updated consultation response and the data in the expert knowledge base to evaluate the quality of the response.

[0055] If the similarity between the updated consultation response and the expert knowledge base data is greater than or equal to the similarity threshold, it indicates that the quality of the response has been significantly improved. At this time, the system can feedback the updated consultation response to the client through the user interface for the user to view. If the similarity between the updated consultation response and the expert knowledge base data is still lower than the similarity threshold, the system will send a prompt message to the user, suggesting that the user further provide detailed information or specific scenarios so that the target model can better understand and process the consultation request. The prompt message can include asking the user whether they can provide more technical details, usage scenarios, or related requirements, etc.

[0056] This embodiment improves the quality of the consultation response by dynamically adjusting the content of the consultation request. It enhances the inference ability of the model and continuously optimizes the performance of the target model through user participation and feedback, providing users with more professional and personalized consultation services.

[0057] To efficiently feedback the user's consultation request, it is necessary to train a target model to support the consultation response. Optionally, in the consultation processing method provided in the embodiment of the present application, the target model is trained in the following way: obtaining expert knowledge data in the field of system operation and maintenance, and preprocessing the expert knowledge data to obtain preprocessed expert knowledge data, where the preprocessing includes at least one of the following: duplicate removal, noise reduction, text cleaning, and word segmentation encoding; converting the preprocessed expert knowledge data into a word vector matrix through a decoder, and adding positional information encoding to the word vector matrix to obtain an updated word vector matrix; determining an initial natural language processing model based on the updated word vector matrix; obtaining historical consultation requests and historical consultation responses in the field of system operation and maintenance, and determining each historical consultation request and the corresponding historical consultation response as a set of training samples to obtain multiple sets of training samples; training the initial natural language processing model based on the multiple sets of training samples to obtain the target model.

[0058] In some examples, expert knowledge data in the field of system operation and maintenance is collected. Such data can come from work order systems, technical documents, forum discussions, Q&A pages, etc., ensuring that the expert knowledge data covers common problems, solutions, operation guides, and best practices in the operation and maintenance process. Duplicate removal refers to eliminating duplicate items in the data to avoid introducing unnecessary biases when training the model. Noise reduction refers to removing useless information in the text, such as advertisements, irrelevant discussions, formatting errors, etc., to ensure data quality. Text cleaning is to standardize the text format, for example, unifying punctuation marks, removing special characters and tags, to ensure the consistency and readability of the text. Word segmentation and encoding is to split the text into words, and then use a pre-trained word embedding model to convert each word into a word vector.

[0059] Use a decoder to convert the preprocessed expert knowledge data into a word vector matrix. The word vector matrix converts each text into a vector representation of a fixed dimension, facilitating model processing. Add position information encoding to the word vector matrix to help the model understand the relative positions of words in the text. The position information encoding allows the model to learn long-distance dependencies in the word sequence.

[0060] Based on the processed word vector matrix and position information, determine an initial natural language processing model. It can be a pre-trained Transformer model. Obtain historical consultation requests and historical consultation responses in the field of system operation and maintenance, and use these data as training samples. Use multiple sets of training samples to fine-tune the initial natural language processing model to meet the specific needs of system operation and maintenance consultations. After training and optimization, the model can better understand and generate responses to system operation and maintenance consultations. At this time, the model is the target model.

[0061] Evaluate the performance of the target model through a sample test set to ensure that it meets the expectations in terms of similarity, accuracy, and relevance. Adjust the model parameters according to the evaluation results, or improve the preprocessing and training steps to continuously optimize the quality of the model's consultation responses.

[0062] Utilize the multi-head attention mechanism. According to the matrix dimension of the word vector matrix, it is split into multiple heads. Multiply the query (Q) vector and the key (K) vector for the matrix, and then multiply by the V vector. The resulting result is used to measure the attention degree to achieve semantic relationship learning.

[0063] Due to the data generated after semantic relationship learning, the discrete values of the data are too large. Use layer normalization technology for layer normalization to obtain the attention weights of each word vector in the word vector matrix, ensuring that the sum of the weights is 1 and the weights are comparable.

[0064] Through the above steps, an output matrix \(A_i\) of masked self-attention is obtained. Then, multiple outputs \(A_i\) are concatenated through the multi-head attention mechanism, and an output \(A\) of the multi-head attention mechanism is calculated. For the output of the attention layer, a non-linear transformation is performed through a fully-connected feed-forward network for learning in the feed-forward neural network layer and recording the learning weights. After learning, due to the large data discreteness, layer normalization is performed again. After \(N\) cycles, a digital queue is generated, which is converted into percentage probabilities through the softmax (normalized exponential) function, and then the word vectors of the output matrix \(A\) are converted into text as the output of the front-end page.

[0065] In this embodiment, by training the target model, it can provide accurate and timely consultation responses, effectively solve the daily consultation needs of operation and maintenance personnel, reduce manual intervention, and improve work efficiency and satisfaction.

[0066] To ensure the accuracy of consultation responses, the target model needs to be updated regularly. Optionally, in the consultation processing method provided in the embodiments of the present application, the method further includes: updating the expert knowledge data in the field of system operation and maintenance every preset period to obtain updated expert knowledge data; updating the word vector matrix of the target model based on the updated expert knowledge data to obtain an updated target model.

[0067] In some instances, by setting a preset period (such as weekly, monthly, or quarterly), within this period, the system automatically or manually collects the latest system operation and maintenance knowledge, including new technical documents, operation guides, common problems and their solutions, industry trends, etc. The newly collected data is integrated with the existing expert knowledge data to obtain updated expert knowledge data.

[0068] The decoder is used to convert the updated expert knowledge data into a word vector matrix, and position information encoding is added to the updated word vector matrix to maintain the correctness of semantic relationships. Based on the updated word vector matrix, the target model is fine-tuned or retrained to ensure that the model can understand and generate consultation responses based on the latest knowledge.

[0069] After the target model is updated, a set of test data is used to evaluate its performance to ensure that the updated data and model parameters improve the accuracy and relevance of consultation responses. If the target model performs poorly in some aspects, for example, the responses to specific questions are not accurate enough, the system can collect feedback from users or manual reviews for further optimizing the target model. The results of evaluation and feedback are used in the next update cycle to continuously iterate the model to adapt to the dynamic changes in the knowledge of the system operation and maintenance field.

[0070] In this embodiment, by regularly updating the expert knowledge data and the target model, it can continuously learn and adapt to new technologies and problems, and maintain the accuracy and timeliness of its consultation responses.

[0071] To evaluate the accuracy of consultation responses, it is necessary to first construct an expert knowledge base for the fields to which the consultation responses may belong. Optionally, in the consultation processing method provided in the embodiments of the present application, before determining the target field to which the consultation request belongs, the method further includes: determining multiple systems and databases applied by the target institution, determining a field based on the operation and maintenance data of each system or database, and obtaining multiple fields; for each field, collecting professional terms, operation and maintenance problems, and problem responses in the field; constructing an expert knowledge base for the field based on the professional terms, operation and maintenance problems, and problem responses.

[0072] In some examples, by reviewing all the systems and databases applied by the target institution, understand the characteristics and operation and maintenance requirements of each system or database. Based on the characteristics, functions, and common operation and maintenance problems of the systems or databases, divide them into different operation and maintenance fields. For example, database operation and maintenance, network operation and maintenance, application system operation and maintenance, security operation and maintenance, etc. can be used as different fields. It is also possible to divide the fields based on different business systems or databases.

[0073] For each field, conduct in-depth research and data collection to construct an expert knowledge base within the field. By sorting out and collecting professional terms within each field, including technical nouns, abbreviations, specific operation commands, etc. Collect common operation and maintenance problems and their descriptions in each field from channels such as historical work orders and forums. Obtain expert responses or solutions to the above operation and maintenance problems, including specific steps, troubleshooting methods, repair codes, etc.

[0074] Sort out the collected professional terms, operation and maintenance problems, and problem responses to ensure the accuracy and consistency of the data. Use database or knowledge graph technology to construct an expert knowledge base, use professional terms and operation and maintenance problems as nodes, and problem responses as edges or attributes between nodes to form a structured knowledge network. Add indexes to the knowledge base to support fast search and query functions, enabling the system to quickly locate expert knowledge related to specific operation and maintenance problems.

[0075] Regularly monitor the operation and maintenance logs of systems and databases, discover new problems or terms, and update the expert knowledge base in a timely manner. Establish a knowledge feedback mechanism to allow operation and maintenance personnel and users to evaluate and give suggestions on the information in the knowledge base, and continuously optimize the knowledge content.

[0076] This embodiment provides a solid data foundation for the consultation processing system by constructing an expert knowledge base for each field. The knowledge base for each field not only contains professional terms and common problems in a specific field, but also covers verified solutions, improving the system's processing ability for complex operation and maintenance consultations and the accuracy of responses.

[0077] It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. And although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0078] Embodiment 2

[0079] The embodiment of the present application also provides a consultation processing device. It should be noted that the consultation processing device of the embodiment of the present application can be used to execute the consultation processing method provided by the embodiment of the present application. The following introduces the consultation processing device provided by the embodiment of the present application.

[0080] According to the embodiment of the present application, there is also provided a device for implementing the above-mentioned consultation processing method. Figure 3 It is a schematic diagram of the consultation processing device provided by the embodiment of the present application, as Figure 3 shown. The device includes:

[0081] A receiving unit 301, configured to receive a consultation request from a user, where the consultation request is a consultation request associated with system operation and maintenance;

[0082] An input unit 302, configured to input the consultation request into a target model to obtain a consultation reply, where the target model is obtained by training with multiple sets of training samples, and each set of training samples includes a historical consultation request and a historical consultation reply;

[0083] A calculation unit 303, configured to determine the target field to which the consultation request belongs, and calculate the similarity between the consultation reply and the data in the expert knowledge base of the target field;

[0084] A first feedback unit 304, configured to, when the similarity is greater than or equal to a similarity threshold, feedback the consultation reply to the client of the user.

[0085] The consultation processing device provided by the embodiment of the present application receives a consultation request from a user through a receiving unit 301, where the consultation request is a consultation request associated with system operation and maintenance; an input unit 302 inputs the consultation request into a target model to obtain a consultation reply, where the target model is obtained by training with multiple groups of training samples, and each group of training samples includes a historical consultation request and a historical consultation reply; a calculation unit 303 determines the target field to which the consultation request belongs and calculates the similarity between the consultation reply and the data in the expert knowledge base of the target field; a first feedback unit 304, when the similarity is greater than or equal to a similarity threshold, feeds back the consultation reply to the user's client. By training the target model, generating a consultation reply to the consultation request by the target model, and feeding back the consultation reply to the user when the similarity between the consultation reply and the data in the expert knowledge base of the target field is greater than or equal to the similarity threshold, the purpose of automatically solving daily consultations in the field of system operation and maintenance is achieved, thereby realizing the technical effect of improving the processing efficiency of consultations, and further solving the technical problem of low consultation processing efficiency in the field of system operation and maintenance.

[0086] Optionally, in the consultation processing device provided by the embodiment of the present application, the calculation unit 303 includes: an extraction module, configured to extract a set of keywords from the consultation request, input the set of keywords into a clustering model, and obtain a clustering result; a first determination module, configured to determine the probability that the consultation request belongs to each field from the clustering result to obtain a set of probabilities; a second determination module, configured to determine the field corresponding to the maximum probability in the set of probabilities as the target field.

[0087] Optionally, in the consultation processing device provided by the embodiment of the present application, the calculation unit 303 includes: a first preprocessing module, configured to preprocess the consultation reply to obtain a preprocessed reply statement, where the preprocessing includes at least one of the following: removing stop words, converting punctuation marks to upper and lower cases, and lemmatization; a processing module, configured to vectorize the reply statement to obtain a target word vector; a calculation module, configured to extract multiple document vectors from the expert knowledge base and calculate the cosine similarity between the target word vector and each document vector to obtain the similarity between the consultation reply and the data in the expert knowledge base of the target field.

[0088] Optionally, in the consultation processing device provided in the embodiments of the present application, the device further includes: a first acquisition unit, configured to acquire a background description of a target field when the similarity is less than a similarity threshold; an addition unit, configured to add the background description to the consultation request to obtain an updated consultation request, and input the updated consultation request into the target model to obtain an updated consultation response; a second feedback unit, configured to, when the similarity between the updated consultation response and the data in the expert knowledge base is greater than or equal to the similarity threshold, feedback the updated consultation response to the user's client; a prompt unit, configured to, when the similarity between the updated consultation response and the data in the expert knowledge base is less than the similarity threshold, send a prompt message, where the prompt message is used to prompt the user to increase the description content of the updated consultation request.

[0089] Optionally, in the consultation processing device provided in the embodiments of the present application, the device further includes: a second acquisition unit, configured to acquire expert knowledge data in the system operation and maintenance field and perform preprocessing on the expert knowledge data to obtain preprocessed expert knowledge data, where the preprocessing includes at least one of the following: duplicate removal, noise reduction, text cleaning, and word segmentation encoding; a conversion unit, configured to convert the preprocessed expert knowledge data into a word vector matrix through a decoder and add position information encoding to the word vector matrix to obtain an updated word vector matrix; a first determination unit, configured to determine an initial natural language processing model based on the updated word vector matrix; a third acquisition unit, configured to acquire historical consultation requests and historical consultation responses in the system operation and maintenance field and determine each historical consultation request and the corresponding historical consultation response as a set of training samples to obtain multiple sets of training samples; a training unit, configured to train the initial natural language processing model based on the multiple sets of training samples to obtain a target model.

[0090] Optionally, in the consultation processing device provided in the embodiments of the present application, the device further includes: a first update unit, configured to update the expert knowledge data in the system operation and maintenance field every preset period to obtain updated expert knowledge data; a second update unit, configured to update the word vector matrix of the target model based on the updated expert knowledge data to obtain an updated target model.

[0091] Optionally, in the consultation processing device provided in the embodiments of the present application, the device further includes: a second determination unit, configured to determine multiple systems and databases applied by a target institution, and determine a field for each system or database based on the operation and maintenance data to obtain multiple fields; an acquisition unit, configured to, for each field, acquire professional terms, operation and maintenance problems, and problem responses of the field; and construct an expert knowledge base of the field based on the professional terms, operation and maintenance problems, and problem responses.

[0092] It should be noted here that the above receiving unit 301, input unit 302, calculation unit 303, and first feedback unit 304 correspond to steps S201 to S204 in Embodiment 1. The functions and application scenarios implemented by the two modules and the corresponding steps are the same, but are not limited to the content disclosed in Embodiment 1 above. It should be noted that the above modules or units may be hardware components or software components stored in a memory (for example, memory 104) and processed by one or more processors (for example, processors 102a, 102b,..., 102n). The above modules may also be part of a device and can run in the computer terminal 10 provided in Embodiment 1.

[0093] Embodiment 3

[0094] An embodiment of the present application may provide an electronic device. Figure 4 It is a structural block diagram of an electronic device according to an embodiment of the present application. As Figure 4 shown, the electronic device may include: one or more ( Figure 4 only one is shown in the figure) processors 402, a memory 404, a storage controller, and a peripheral interface. Among them, the peripheral interface is connected to a radio frequency module, an audio module, and a display.

[0095] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the methods and devices in the embodiments of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, implements the above methods. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memories, or other non-volatile solid-state memories. In some instances, the memory may further include a memory remotely set relative to the processor, and these remote memories can be connected to the terminal through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0096] The processor can call the information and application programs stored in the memory through a transmission device to execute the following steps: receiving a consultation request from a user, where the consultation request is a consultation request associated with system operation and maintenance; inputting the consultation request into a target model to obtain a consultation response, where the target model is obtained by training with multiple sets of training samples, and each set of training samples includes a historical consultation request and a historical consultation response; determining the target field to which the consultation request belongs, and calculating the similarity between the consultation response and the data in the expert knowledge base of the target field; in the case where the similarity is greater than or equal to a similarity threshold, feeding back the consultation response to the user's client.

[0097] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: extract a set of keywords from the consultation request, input the set of keywords into the clustering model to obtain a clustering result; determine the probability that the consultation request belongs to each field from the clustering result to obtain a set of probabilities; and determine the field corresponding to the maximum probability in the set of probabilities as the target field.

[0098] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: preprocess the consultation reply to obtain a preprocessed reply statement, where the preprocessing includes at least one of the following: removing stop words, converting punctuation marks, converting case, and lemmatization; vectorize the reply statement to obtain a target word vector; extract multiple document vectors from the expert knowledge base, and calculate the cosine similarity between the target word vector and each document vector to obtain the similarity between the consultation reply and the data in the expert knowledge base of the target field.

[0099] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: when the similarity is less than the similarity threshold, obtain the background description of the target field; add the background description to the consultation request to obtain an updated consultation request, input the updated consultation request into the target model to obtain an updated consultation reply; when the similarity between the updated consultation reply and the data in the expert knowledge base is greater than or equal to the similarity threshold, feedback the updated consultation reply to the user's client; when the similarity between the updated consultation reply and the data in the expert knowledge base is less than the similarity threshold, send a prompt message, where the prompt message is used to prompt the user to increase the description content of the updated consultation request.

[0100] The processor can also call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain the expert knowledge data in the system operation and maintenance field, and preprocess the expert knowledge data to obtain preprocessed expert knowledge data, where the preprocessing includes at least one of the following: deduplication, noise reduction, text cleaning, and tokenization and encoding; convert the preprocessed expert knowledge data into a word vector matrix through a decoder, and add positional information encoding to the word vector matrix to obtain an updated word vector matrix; determine an initial natural language processing model based on the updated word vector matrix; obtain the historical consultation requests and historical consultation replies in the system operation and maintenance field, and determine each historical consultation request and the corresponding historical consultation reply as a set of training samples to obtain multiple sets of training samples; and train the initial natural language processing model based on the multiple sets of training samples to obtain a target model.

[0101] The processor can also call the information and application programs stored in the memory through a transmission device to perform the following steps: update the expert knowledge data in the field of system operation and maintenance at preset intervals to obtain updated expert knowledge data; update the word vector matrix of the target model based on the updated expert knowledge data to obtain an updated target model.

[0102] The processor can also call the information and application programs stored in the memory through a transmission device to perform the following steps: determine multiple systems and databases applied by the target organization, determine a field based on the operation and maintenance data of each system or database to obtain multiple fields; for each field, collect professional terms, operation and maintenance problems, and problem responses in the field; construct an expert knowledge base for the field based on the professional terms, operation and maintenance problems, and problem responses.

[0103] An embodiment of the present application provides a solution for receiving a user's consultation request, where the consultation request is a consultation request related to system operation and maintenance; inputting the consultation request into a target model to obtain a consultation response, where the target model is trained by multiple sets of training samples, and each set of training samples includes a historical consultation request and a historical consultation response; determining the target field to which the consultation request belongs, and calculating the similarity between the consultation response and the data in the expert knowledge base of the target field; and in the case where the similarity is greater than or equal to a similarity threshold, feedbacking the consultation response to the user's client. By training the target model, generating a consultation response to the consultation request by the target model, and feedbacking the consultation response to the user when the similarity between the consultation response and the data in the expert knowledge base of the target field is greater than or equal to the similarity threshold, the purpose of automatically solving the daily consultations in the field of system operation and maintenance is achieved, thereby realizing the technical effect of improving the processing efficiency of consultations, and further solving the technical problem of low consultation processing efficiency in the field of system operation and maintenance.

[0104] Those of ordinary skill in the art can understand that Figure 4 the structure shown is only schematic, and the electronic device can also be a terminal device such as a smart phone, a tablet computer, a personal digital assistant, and a Mobile Internet Device (MID), a PAD, etc. Figure 4 It does not limit the structure of the above-mentioned electronic device. For example, the electronic device may further include more or fewer components (such as a network interface, a display device, etc.) than those shown Figure 4 in the figure, or have a different configuration from that shown Figure 4 in the figure.

[0105] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by a program instructing the hardware related to the terminal device. The program can be stored in a computer-readable storage medium, and the storage medium can include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, an optical disk, etc.

[0106] Embodiment 4

[0107] An embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the above storage medium can be used to store the program code executed by the consultation processing method provided in the first embodiment above.

[0108] Optionally, in this embodiment, the above storage medium can be located in any one of the computer terminals in the computer terminal group in the computer network, or in any one of the mobile terminals in the mobile terminal group.

[0109] The present application also provides a computer program product, which is adapted to execute a program for the steps of the consultation processing method when executed on a data processing device.

[0110] The serial numbers of the above embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments.

[0111] In the above embodiments of the present application, the descriptions of the various embodiments have their own emphases. For the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0112] In the several embodiments provided by the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point, the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces, and the indirect coupling or communication connection of the units or modules can be in an electrical or other form.

[0113] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or they can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0114] In addition, in each embodiment of the present application, the functional units may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional unit.

[0115] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, may be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, read-only memories (ROMs), random access memories (RAMs), mobile hard disks, magnetic disks, or optical discs that can store program codes.

[0116] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present application, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present application.

Claims

1. A consultation processing method, characterized in that: include: Receiving a consultation request from a user, wherein the consultation request is a consultation request associated with system operation and maintenance; Inputting the consultation request into a target model to obtain a consultation response, wherein the target model is trained by multiple sets of training samples, each set of training samples including historical consultation requests and historical consultation responses; Determine the target field to which the consultation request belongs, and calculate the similarity between the consultation response and the data of the expert knowledge base of the target field; When the similarity is greater than or equal to a similarity threshold, the consultation reply is fed back to the user's client.

2. The method according to claim 1, characterized in that: The target areas for the consultation request include: Extracting a group of keywords from the consultation request, and inputting the group of keywords into a clustering model to obtain a clustering result; Determine the probability that the consultation request belongs to each field from the clustering results to obtain a set of probabilities; The domain corresponding to the maximum probability in the set of probabilities is determined as the target domain.

3. The method according to claim 1, characterized in that: Calculating the similarity between the consultation response and the data of the expert knowledge base in the target field includes: Preprocessing the consultation reply to obtain a preprocessed reply statement, wherein the preprocessing includes at least one of the following: removing stop words, converting punctuation marks to upper and lower case, and restoring morphology; Vectorize the reply sentence to obtain a target word vector; A plurality of document vectors are extracted from the expert knowledge base, and the cosine similarity between the target word vector and each document vector is calculated to obtain the similarity between the consultation response and the data of the expert knowledge base in the target field.

4. The method according to claim 1, characterized in that After calculating the similarity between the consultation response and the data in the expert knowledge base of the target field, the method further includes: When the similarity is less than the similarity threshold, obtaining a background description of the target domain; Adding the background description to the consultation request to obtain an updated consultation request, inputting the updated consultation request into the target model to obtain an updated consultation response; When the similarity between the updated consultation reply and the data of the expert knowledge base is greater than or equal to the similarity threshold, feeding back the updated consultation reply to the client of the user; When the similarity between the updated consultation reply and the data of the expert knowledge base is less than the similarity threshold, a prompt message is issued, wherein the prompt message is used to prompt the user to increase the description content of the updated consultation request.

5. The method according to claim 1, characterized in that The target model is trained in the following way: Acquire expert knowledge data in the field of system operation and maintenance, and preprocess the expert knowledge data to obtain preprocessed expert knowledge data, wherein the preprocessing includes at least one of the following: deduplication, noise reduction, text cleaning, and word segmentation encoding; The preprocessed expert knowledge data is converted into a word vector matrix by a decoder, and position information encoding is added to the word vector matrix to obtain an updated word vector matrix; Determining an initial natural language processing model based on the updated word vector matrix; Obtaining historical consultation requests and historical consultation replies in the field of system operation and maintenance, and determining each historical consultation request and corresponding historical consultation reply as a set of training samples, thereby obtaining multiple sets of training samples; The initial natural language processing model is trained based on the multiple groups of training samples to obtain the target model.

6. The method according to claim 5, characterized in that The method further comprises: updating the expert knowledge data in the field of system operation and maintenance at preset intervals to obtain updated expert knowledge data; The word vector matrix of the target model is updated based on the updated expert knowledge data to obtain an updated target model.

7. The method according to claim 1, characterized in that Before determining the target field to which the consultation request belongs, the method further includes: Identify multiple systems and databases used by the target organization, determine a field based on the operation and maintenance data of each system or database, and obtain multiple fields; For each field, collect the professional terms, operation and maintenance issues, and responses to the questions in the field; An expert knowledge base in the field is constructed based on the professional terms, the operation and maintenance issues, and the responses to the issues.

8. A consultation processing device, characterized in that: include: A receiving unit, configured to receive a consultation request from a user, wherein the consultation request is a consultation request associated with system operation and maintenance; An input unit, used for inputting the consultation request into a target model to obtain a consultation reply, wherein the target model is trained by multiple sets of training samples, each set of training samples including historical consultation requests and historical consultation replies; A calculation unit, used to determine the target field to which the consultation request belongs, and calculate the similarity between the consultation response and the data of the expert knowledge base of the target field; The first feedback unit is used to feed back the consultation reply to the user's client when the similarity is greater than or equal to a similarity threshold.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is executed, the device where the computer-readable storage medium is located is controlled to execute the consultation processing method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: A memory storing an executable program; A processor is used to run the program, wherein the program, when running, executes the consultation processing method described in any one of claims 1 to 7.

11. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the consultation processing method described in any one of claims 1 to 7 are implemented.

Images