Intelligent business consultation method and device for mobile bank

Abstract

The invention provides an intelligent business consultation method and device for a mobile bank, and relates to the technical field of data processing, and the method comprises the steps: receiving consultation request release, obtaining a response instruction, carrying out priority analysis, semantic recognition and sensitivity analysis, then carrying out consultation reply analysis, encrypting a reply result, and sending the reply result to a user. The method can solve the problems that mass data is difficult to process efficiently in the prior art, variable requirements cannot be met flexibly due to rigid priority setting, reply content is deviated or inaccurate due to insufficient understanding accuracy of semantic intentions of users, a necessary protection mechanism is lacked to ensure data security when sensitive information is processed, and the user experience is poor. By comprehensively optimizing concurrent processing, user priority, semantic understanding, sensitive information protection and data transmission security, the efficiency, security and user experience of mobile bank services are remarkably improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to an intelligent business consulting method and device for mobile banking. Background Art

[0002] As the number of customers continues to increase, the banking industry is facing problems such as rising labor costs, limited service efficiency, and difficulty in breaking through service quality. Changes in customer needs have also promoted the development of intelligent business consulting methods. Customers are paying more and more attention to the convenience and speed of services, and they expect to be able to obtain bank services and support anytime and anywhere. At the same time, customers' personalized needs for services are also increasing, and they hope to get more intimate and professional services.

[0003] At present, the existing customer service methods, such as telephone communication and PC webpage online customer service, can no longer meet the growing customer needs. Therefore, the banking industry needs to seek a new service model to improve customer experience and service efficiency.

[0004] In summary, existing technologies are unable to efficiently process massive amounts of data, have rigid priority settings that cannot flexibly respond to changing demands, lack accurate understanding of user semantic intent that leads to deviated or inaccurate responses, and lack the necessary protection mechanisms to ensure data security when processing sensitive information. Summary of the invention

[0005] The purpose of this application is to provide an intelligent business consulting method and device for mobile banking, so as to solve the problems that the existing technology is difficult to efficiently process massive data, the priority setting is rigid and cannot flexibly respond to changing needs, the understanding of user semantic intent is not accurate enough to cause deviation or inaccuracy in the reply content, and there is a lack of necessary protection mechanisms to ensure data security when processing sensitive information.

[0006] In view of the above problems, the present application provides an intelligent business consulting method and device for mobile banking.

[0007] In a first aspect, the present application provides an intelligent business consulting method for mobile banking, the method being implemented by an intelligent business consulting device for mobile banking, wherein the method comprises: receiving a concurrent consulting request distribution according to a human-machine interface of a target mobile bank, wherein the concurrent consulting request distribution comprises a plurality of user consulting request information corresponding to a plurality of users; performing load verification on the concurrent consulting request distribution according to a concurrent consulting load constraint to obtain a concurrent request response instruction; performing business consulting priority analysis on the plurality of users based on the concurrent request response instruction to obtain a user business consulting priority constraint sequence, and classifying the concurrent consulting request according to the user business consulting priority constraint sequence. The f-th user consultation request information corresponding to the f-th user is generated by traversing and extracting the f-th user, wherein f is a positive integer, f belongs to F, and F is the total number of users of the multiple users; the multi-business consultation semantic analysis branch is activated to perform semantic recognition on the f-th user consultation request information to generate the f-th consultation semantic feature data; according to the consultation sensitivity evaluation branch, the f-th consultation semantic feature data is subjected to deep sensitive analysis to generate the f-th consultation activation signal; according to the f-th consultation activation signal, the business consultation reply channel is activated to perform consultation reply analysis on the f-th consultation semantic feature data to generate the f-th consultation reply result; according to the human-computer linkage interface, the f-th consultation reply result is encrypted and sent to the f-th user.

[0008] In a second aspect, the present application also provides an intelligent business consulting device for mobile banking, which is used to execute the intelligent business consulting method for mobile banking as described in the first aspect, wherein the device includes: a request information acquisition module, the request information acquisition module is used to receive a concurrent consultation request distribution according to the human-computer linkage interface of the target mobile bank, wherein the concurrent consultation request distribution includes multiple user consultation request information corresponding to multiple users; a response instruction acquisition module, the response instruction acquisition module is used to perform load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint, and obtain a concurrent request response instruction; a constraint sequence acquisition module, the constraint sequence acquisition module is used to perform business consultation priority analysis on the multiple users based on the concurrent request response instruction, obtain a user business consultation priority constraint sequence, and traverse the concurrent consultation request distribution according to the user business consultation priority constraint sequence. History extraction, generating the fth user consultation request information corresponding to the fth user, wherein f is a positive integer, f belongs to F, and F is the total number of users of the multiple users; a semantic recognition module, the semantic recognition module is used to activate the multi-business consultation semantic analysis branch to perform semantic recognition on the fth user consultation request information, and generate the fth consultation semantic feature data; an activation signal generation module, the activation signal generation module is used to perform a deep sensitive analysis on the fth consultation semantic feature data according to the consultation sensitive evaluation branch, and generate the fth consultation activation signal; a reply result generation module, the reply result generation module is used to activate the business consultation reply channel according to the fth consultation activation signal, perform consultation reply analysis on the fth consultation semantic feature data, and generate the fth consultation reply result; a result sending module, the result sending module is used to encrypt the fth consultation reply result and send it to the fth user according to the human-computer linkage interface.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages: By receiving the concurrent consultation request distribution according to the human-machine interaction interface of the target mobile bank, wherein the concurrent consultation request distribution includes a plurality of user consultation request information corresponding to a plurality of users; performing load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint to obtain a concurrent request response instruction; based on the concurrent request response instruction, performing business consultation priority analysis on the plurality of users to obtain a user business consultation priority constraint sequence, and traversing and extracting the concurrent consultation request distribution according to the user business consultation priority constraint sequence to generate the fth user consultation request information corresponding to the fth user, wherein f is a positive integer, f belongs to F, and F is the total number of users of the plurality of users; activating the multi-business consultation semantic analysis branch to perform semantic recognition on the fth user consultation request information to generate the fth consultation semantic feature data; according to the consultation The inquiry sensitivity evaluation branch performs deep sensitive analysis on the fth inquiry semantic feature data to generate an fth inquiry activation signal; based on the fth inquiry activation signal, activates the business inquiry reply channel to perform inquiry reply analysis on the fth inquiry semantic feature data to generate an fth inquiry reply result; according to the human-machine linkage interface, encrypts the fth inquiry reply result and sends it to the fth user, which effectively solves the problems that the existing technology is difficult to efficiently process massive data, the priority setting is rigid and cannot flexibly respond to changing needs, the understanding of user semantic intentions is not accurate enough to cause deviation or inaccuracy in reply content, and there is a lack of necessary protection mechanisms to ensure data security when processing sensitive information. By comprehensively optimizing concurrent processing, user priority, semantic understanding, sensitive information protection and data transmission security, the efficiency, security and user experience of mobile banking services are significantly improved.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0012] Figure 1 A flowchart of the intelligent business consulting method for mobile banking used in this application; Figure 2 This is a schematic diagram of the structure of the intelligent business consulting device used for mobile banking in this application.

[0013] Description of reference numerals: Request information acquisition module 11, response instruction acquisition module 12, constraint sequence acquisition module 13, semantic recognition module 14, activation signal generation module 15, reply result generation module 16, result sending module 17. DETAILED DESCRIPTION

[0014] The present application provides an intelligent business consulting method and device for mobile banking, which solves the problems that the prior art is difficult to efficiently process massive data, the priority setting is rigid and cannot flexibly respond to changing needs, the understanding of user semantic intent is not accurate enough resulting in deviation or inaccurate response content, and there is a lack of necessary protection mechanisms to ensure data security when processing sensitive information. By comprehensively optimizing concurrent processing, user priority, semantic understanding, sensitive information protection and data transmission security, the efficiency, security and user experience of mobile banking services are significantly improved.

[0015] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0016] Embodiment 1 Please see attached Figure 1 The present application provides an intelligent business consulting method for mobile banking, wherein the method is applied to an intelligent business consulting device for mobile banking, and the method specifically comprises the following steps: Step 1: receiving a concurrent consultation request distribution according to a human-machine interaction interface of a target mobile bank, wherein the concurrent consultation request distribution includes a plurality of user consultation request information corresponding to a plurality of users; Specifically, build a server architecture that supports high concurrency, such as a distributed system, to handle a large number of requests that arrive at the same time. Use load balancing technology to distribute requests to different servers to ensure that the processing capacity of each server is fully utilized. Implement request queue management to sort and prioritize incoming requests to ensure that important or urgent requests are processed first. Provide consultation portals on the human-computer interaction interface, such as chat windows, help buttons, etc. When a user submits a consultation request, automatically record user information and request content, including timestamp, user ID, request type, etc. Use back-end services to monitor and receive these requests in real time to ensure that every user consultation is processed without omission. Distribute it to the corresponding processing module according to the type and content of the request, such as the self-service module, manual customer service module, etc.

[0017] Step 2: Perform load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint to obtain a concurrent request response instruction; Specifically, key performance indicators such as the maximum number of concurrent consultation requests, response time, and resource utilization that can be tolerated are clearly defined. These can be set based on historical data and business requirements to ensure that the system does not exceed its processing capacity during normal operation. Use load testing tools or custom scripts to simulate scenarios where multiple users initiate consultation requests at the same time. The distribution of requests should be as close to the actual business scenario as possible, including the type, frequency, and size of requests. Send simulated concurrent consultation requests to the human-computer interaction interface and monitor the system's response. Collect key performance indicator data such as response time, throughput, error rate, and resource utilization. Compare the collected data with predefined load constraints to evaluate whether the system meets performance requirements. Identify performance bottlenecks and potential problems, such as request processing delays, resource contention, and memory leaks. Make necessary adjustments and optimizations based on the verification results. This includes increasing server resources, optimizing code logic, and improving database queries. Repeat the load verification process until the system performance meets the predefined load constraints. Generate concurrent request response instructions based on the verification results and the current state of the system. These instructions can guide the system on how to handle concurrent consultation requests in actual operation. For example, when the system load is light, the system can be instructed to prioritize high-priority or specific types of requests; when the system load is heavy, the system can be instructed to initiate current limiting, queuing, or degradation strategies.

[0018] Step 3: Based on the concurrent request response instruction, perform business consultation priority analysis on the multiple users to obtain a user business consultation priority constraint sequence, and traverse and extract the concurrent consultation request distribution according to the user business consultation priority constraint sequence to generate the fth user consultation request information corresponding to the fth user, where f is a positive integer, f belongs to F, and F is the total number of users of the multiple users; Specifically, concurrent request response instructions are received, which include information such as the priority of processing requests, resource allocation strategies, and the current load status of the system. The received concurrent request response instructions are parsed to clarify the priority rules for user business consultations. The priority rules are based on multiple factors such as user level, request type, request urgency, and service level agreement. The consultation request information of multiple users currently to be processed is obtained from the request queue or database. This information includes user ID, request time, request content, request type, etc. According to the parsed priority rules, the obtained multiple user consultation request information is sorted. The result of the sorting is the user business consultation priority constraint sequence. According to the order of the user business consultation priority constraint sequence, the consultation request of each user is traversed and extracted. For the fth user in the sequence, f is a positive integer, and f belongs to F, and F is the total number of multiple users, the consultation request information of the user is extracted, that is, the fth user consultation request information corresponding to the fth user. According to the extracted fth user consultation request information, the corresponding business processing is performed.

[0019] Step 4: activating the multi-business consultation semantic analysis branch to perform semantic recognition on the f-th user consultation request information to generate the f-th consultation semantic feature data; Specifically, the semantic parsing branch is trained and has the ability to semantically recognize user consultation requests. This branch is a deep learning model based on natural language processing, such as recurrent neural networks, long short-term memory networks, etc. The extracted f-th user consultation request information is passed as input data to the multi-business consultation semantic parsing branch. The input data includes the text content of the user consultation, user ID, request time, etc. The input f-th user consultation request information is preprocessed, including removing irrelevant characters, punctuation marks, stop words, etc. The word segmentation operation is performed to divide the continuous text into independent vocabulary units. It is also necessary to perform word vectorization to convert the text into a numerical form that the model can understand. The multi-business consultation semantic parsing branch is used to semantically recognize the preprocessed f-th user consultation request information. The model analyzes the vocabulary, phrases, sentence structure and contextual information in the text to understand the user's consultation intention and needs. According to the results of semantic recognition, the semantic features related to the f-th user consultation request are extracted. These features include the topic of the user consultation, intent classification, key entities, emotional tendency, etc. These features are organized in a structured form to generate the f-th consultation semantic feature data.

[0020] Step 5: performing a deep sensitivity analysis on the fth consultation semantic feature data according to the consultation sensitivity evaluation branch to generate the fth consultation activation signal; Specifically, ensure that the f-th consultation semantic feature data generated by semantic recognition is available. This data should contain semantic features such as key information, intent, and entities of the user's consultation. The consultation sensitivity evaluation branch is a model or algorithm specifically used to analyze the sensitivity of consultation semantic feature data. This branch has been properly trained and has the ability to perform sensitive evaluation on consultation data. The f-th consultation semantic feature data is passed as input to the consultation sensitivity evaluation branch. The consultation sensitivity evaluation branch will conduct a deep analysis of the input data to evaluate whether it contains sensitive information. Sensitive information involves personal privacy, commercial secrets, compliance with laws and regulations, etc. The branch will use preset rules, models, or algorithms to identify and evaluate the sensitivity of this information. Based on the results of the deep sensitivity analysis, the consultation sensitivity evaluation branch will generate a sensitivity evaluation report or score. This report or score reflects the degree and type of sensitive information in the f-th consultation semantic feature data. Based on the sensitivity evaluation results, the system will determine whether it is necessary to generate the f-th consultation activation signal. If the sensitive information contained in the data exceeds the preset threshold or meets specific activation conditions, the system will generate an activation signal. An activation signal is a flag, instruction, or message used to trigger subsequent processing.

[0021] Step 6: according to the fth consultation activation signal, activate the business consultation reply channel to perform consultation reply analysis on the fth consultation semantic feature data to generate the fth consultation reply result; Specifically, the fth consultation activation signal is received, which indicates that the consultation request of the fth user needs to be analyzed for a reply. It is ensured that only consultation requests that have passed the sensitivity assessment and meet the requirements will enter the reply analysis stage. Based on the received activation signal, the corresponding business consultation reply channel is activated. This channel is a module used to process user consultation requests and generate replies. The fth consultation semantic feature data generated by semantic recognition is used as input and passed to the business consultation reply channel. These data contain key information of user consultation, such as consultation topics, intentions, key entities, etc., which are important basis for generating replies. In the business consultation reply channel, the fth consultation semantic feature data is deeply analyzed. The analysis process includes understanding the user's consultation intention, retrieving relevant information, applying business rules or logic for reasoning, etc. Based on the results of the consultation reply analysis, the fth consultation reply result is generated. The reply result is a piece of text, an instruction, a link or other forms of information. When generating a reply, multiple factors are considered comprehensively, such as user preferences, historical interaction records, business goals, etc., to ensure the accuracy and satisfaction of the reply.

[0022] Step seven: Encrypt the fth consultation reply result and send it to the fth user according to the human-computer linkage interface.

[0023] Specifically, prepare the corresponding encryption key, which will be used to encrypt and decrypt data to ensure the confidentiality of the data. Obtain the fth consultation response result from the business consultation response channel or related processing module. Ensure that the received response result is accurate and complete. Use encryption tools and keys to encrypt the fth consultation response result. Encryption can be performed using symmetric encryption, asymmetric encryption, or hybrid encryption, depending on security requirements and configuration. The encrypted data will become unreadable, and only users holding the corresponding keys can decrypt and view the original content. Pass the encrypted fth consultation response result to the human-computer linkage interface. The human-computer linkage interface is a platform for interacting with users, which is a web page, mobile application or other form of client. The interface will be responsible for sending the encrypted data to the fth user in a secure manner. The fth user receives the encrypted consultation response result through the human-computer linkage interface. The user needs to use the assigned or shared key to decrypt the received data. After successful decryption, the user will be able to view and understand the original consultation response content.

[0024] Furthermore, step 2 of this application also includes: Generating concurrent request capacity according to the concurrent consultation request distribution; Calculating a concurrent consultation load factor based on the concurrent request constraint capacity and the concurrent request capacity; Determining whether the concurrent consultation load factor satisfies the concurrent consultation load constraint; If the concurrent consultation load factor does not satisfy the concurrent consultation load constraint, generating a concurrent request warning instruction; If the concurrent consultation load factor satisfies the concurrent consultation load constraint, the concurrent request response instruction is obtained.

[0025] Specifically, the concurrent consultation request distribution refers to the number and distribution of consultation requests initiated concurrently by users within a period of time. The concurrent request capacity can be calculated by analyzing the distribution of concurrent consultation requests. This capacity represents the maximum number of requests that can be processed at the same time. The concurrent request constraint capacity is the maximum number of requests that can be processed simultaneously, which is preset or determined based on resource constraints. The concurrent consultation load factor is an indicator used to measure the relationship between the current concurrent request capacity and the constraint capacity. It can be calculated by the following formula: concurrent consultation load factor = (current concurrent request capacity / concurrent request constraint capacity) * 100%. The concurrent consultation load constraint is a threshold value of the set load factor, which is used to determine whether the system is overloaded. The calculated concurrent consultation load factor is compared with the preset load constraint. If the load factor exceeds the preset load constraint, it means that there is a risk of overload. At this time, a concurrent request warning instruction is generated to notify relevant managers or automatically trigger the load balancing mechanism. If the load factor is within the preset load constraint range, it means that the concurrent requests can be processed normally at present. At this time, continue to execute or generate concurrent request response instructions to process the user's consultation request.

[0026] Furthermore, step 4 of this application also includes: The multiple business consultation semantic analysis branch includes a request voice noise reduction branch, a request voice recognition branch and a request semantic analysis branch; When the f-th user consultation request information is a voice signal, performing noise cleaning on the f-th user consultation request information according to the request voice noise reduction branch to generate f-th cleaned voice request information; Inputting the fth cleaning voice request information into the request voice recognition branch to obtain the fth voice recognition result; The fth speech recognition result is subjected to deep semantic analysis according to the request semantic analysis branch to obtain the fth consultation semantic feature data, wherein the request semantic analysis branch includes a consultation word semantic analysis layer, a consultation explicit semantic analysis layer and a consultation implicit semantic analysis layer.

[0027] Specifically, when the fth user consultation request information is input in the form of a voice signal, it will first pass through the request voice noise reduction branch. The main task of this branch is to remove background noise, interfering sounds or other non-voice components in the voice signal to improve the clarity and quality of the voice signal. Through noise reduction processing, the fth clean voice request information can be generated, that is, a pure voice signal with noise removed. The fth clean voice request information will be passed to the request voice recognition branch. This branch uses voice recognition technology to convert voice signals into text data. The voice recognition process includes the collaborative work of components such as acoustic models, language models, and dictionaries. Through voice recognition, the fth voice recognition result, that is, the text representation of the user consultation request, can be obtained. Once the fth voice recognition result is obtained, this text data will be input into the request semantic analysis branch for deep semantic analysis. The consultation word semantic analysis layer is responsible for semantic understanding and annotation of vocabulary in the text. This includes tasks such as part-of-speech tagging, named entity recognition, and word sense disambiguation to identify key information and entities in the text. The consultation explicit semantic analysis layer focuses on the semantic relationships clearly expressed in the text. This layer uses dependency syntactic analysis, semantic role labeling and other technologies to capture explicit semantic structures such as subject-verb-object relationships, verb tenses, and modification relationships in sentences. The consultation implicit semantic analysis layer mines semantic information that is not explicitly expressed but can be inferred in the text. It includes advanced semantic processing tasks such as contextual reasoning, coreference resolution, and sentiment analysis to reveal the user's potential intentions and attitudes. Through the in-depth analysis of these three levels, the request semantic analysis branch can fully understand the semantic content of the user's consultation request and generate the fth consultation semantic feature data, which is the basis for subsequent business processing and response.

[0028] Furthermore, step five of this application also includes: The consultation sensitivity evaluation branch includes a sensitive business consultation library, a consultation sensitivity identification branch and a consultation sensitivity judgment branch; According to the consultation sensitivity identification branch, correlation analysis is performed on the fth consultation semantic feature data and each sensitive business consultation feature data in the sensitive business consultation library to generate a plurality of feature sensitive consultation coefficients; Input the plurality of feature-sensitive consulting coefficients into the consulting sensitivity judgment branch, and determine whether the plurality of feature-sensitive consulting coefficients are less than a sensitive consulting threshold in the consulting sensitivity judgment branch; If any one of the plurality of characteristic sensitive consultation coefficients is greater than / equal to the sensitive consultation threshold, generating an f-th consultation taboo signal; If the plurality of characteristic sensitive consultation coefficients are all smaller than the sensitive consultation threshold, the fth consultation activation signal is obtained.

[0029] Specifically, the sensitive business consultation library is a collection of multiple sensitive business consultation feature data. These feature data include keywords, phrases, sentences or business scenarios, etc., which are used to identify and recognize consultations involving sensitive information. When the fth consultation semantic feature data is received, the consultation sensitivity identification branch will start working. This branch uses association analysis technology to compare and match the fth consultation semantic feature data with each sensitive business consultation feature data in the sensitive business consultation library. The purpose of association analysis is to calculate the similarity or correlation between the consultation semantics and the sensitive features, thereby generating multiple feature sensitive consultation coefficients. These coefficients reflect the sensitivity of the consultation content on different sensitive features. Once multiple feature sensitive consultation coefficients are generated, these coefficients will be input into the consultation sensitivity judgment branch. A sensitive consultation threshold is set inside this branch to determine whether the consultation content reaches the sensitive level. This threshold can be set according to business requirements and security standards. The judgment branch compares the size relationship between each feature sensitive consultation coefficient and the sensitive consultation threshold. If any of the multiple feature sensitive consultation coefficients is greater than or equal to the sensitive consultation threshold, it means that the consultation content involves sensitive information. At this time, the fth consultation taboo signal will be generated to prevent or restrict further processing of the consultation. If all feature sensitive consultation coefficients are less than the sensitive consultation threshold, it means that the consultation content is not sensitive. The system will generate the fth consultation activation signal, allowing subsequent business processing and analysis of the consultation.

[0030] Furthermore, step six of this application also includes: Extracting the fth consulting service attribute according to the fth consulting semantic feature data; Based on the fth consultation activation signal, activating the matching business consultation reply branch corresponding to the fth consultation business attribute in the business consultation reply channel; Perform consultation reply analysis on the fth consultation semantic feature data according to each business consultation responder in the matching business consultation reply branch to obtain multiple initial consultation reply results; Data fusion is performed based on the multiple initial consultation response results to obtain the fth consultation response result.

[0031] Specifically, first, the semantic feature data of the fth consultation is deeply analyzed to extract the business attributes contained in the consultation. Business attributes include key information such as consultation type, such as loan consultation, deposit consultation, financial consultation, etc., customer identity, such as individual customers, corporate customers, etc., and consultation urgency. After confirming the receipt of the fth consultation activation signal, the matching business consultation reply branch corresponding to the fth consultation business attribute in the business consultation reply channel is activated. Each business consultation reply branch is optimized for a specific business type or scenario to ensure that the most relevant and accurate consultation reply can be provided. The activated matching business consultation reply branch contains multiple business consultation responders, which are built based on rules, templates, machine learning models, etc. Each responder will independently analyze the semantic feature data of the fth consultation and generate the corresponding initial consultation reply result. This process includes multiple technologies such as information retrieval, logical reasoning, and natural language generation to ensure the accuracy and completeness of the reply. The initial consultation reply results generated by all business consultation responders are collected and data fusion processing is performed. Data fusion includes operations such as result deduplication, information integration, and priority sorting to form a comprehensive, consistent, and high-quality final consultation reply result. Finally, the fused f-th consultation response result will be output to the user to meet his or her consultation needs.

[0032] Furthermore, the present application also includes: Performing a global consultation request reply sample search according to the target mobile bank to obtain a business consultation request reply record set, and performing data cleaning on the business consultation request reply record set to generate a business consultation request reply data set; Classifying consulting business attributes according to the business consulting request reply data set to obtain multiple consulting request reply sample partitions corresponding to multiple sample consulting business attributes; Based on the neural network, a business consultation reply channel architecture is constructed, wherein the business consultation reply channel architecture includes an input layer, a hidden layer and an output layer; Based on the multiple consultation request response sample partitions, training multiple business consultation response branches; The multiple business consultation reply branches are embedded into the hidden layer to generate a consultation reply analysis layer, and the input layer and the output layer are combined to generate the business consultation reply channel.

[0033] Specifically, for the target mobile bank, a comprehensive consultation request and response sample retrieval is performed. This includes collecting relevant consultation requests and corresponding responses from historical records, databases or log files. All the retrieved consultation requests and responses are organized into a record set. The record set is cleaned to remove duplicate, invalid or low-quality records to ensure the accuracy and completeness of the data. The cleaned data is called the business consultation request response data set. The cleaned data set is classified by consultation business attributes. This means that the data is divided into different categories according to the type, content, urgency and other attributes of the consultation. Based on the classification results, the data set is divided into multiple sample partitions, each of which corresponds to a specific consultation business attribute. Neural networks are used as the basis for building the business consultation response channel architecture. Neural networks have strong learning and representation capabilities and are suitable for processing complex natural language processing tasks. The business consultation response channel architecture includes an input layer, a hidden layer and an output layer. The input layer is responsible for receiving the user's consultation request, the hidden layer performs complex processing and analysis, and the output layer generates the final response result. Using the multiple consultation request response sample partitions previously divided, multiple business consultation response branches are trained separately. Each branch focuses on processing consultation requests with specific business attributes. Through continuous training and optimization, we ensure that each business consultation response branch can provide accurate and efficient responses. The trained multiple business consultation response branches are embedded into the hidden layer of the neural network to form a consultation response analysis layer. This layer is responsible for selecting the appropriate business consultation response branch for processing based on the input user consultation request. The consultation response analysis layer is combined with the input layer and the output layer to generate a complete business consultation response channel. The input layer receives the user consultation request, and after processing and analysis by the consultation response analysis layer, the output layer generates and outputs the final consultation response result.

[0034] Furthermore, the present application also includes: Extracting a first consultation request reply sample partition according to the plurality of consultation request reply sample partitions; Determine whether the first sample partition data volume of the first consultation request reply sample partition is less than the preset partition data volume; If the data volume of the first sample partition is less than the data volume of the preset partition, searching for samples with the same consulting business attributes for the first consultation request reply sample partition to generate a first supplementary data set of samples with the same attributes; Using the first same-attribute request reply sample supplementary data set to expand the first consultation request reply sample partition, to generate a first expanded request reply sample partition; According to multiple feature machine learning algorithms, the first expansion request reply sample partition is trained and tested respectively to generate a first business consultation reply branch, and the first business consultation reply branch is added to the multiple business consultation reply branches, wherein the first business consultation reply branch includes multiple first attribute business consultation responders that meet the consultation reply accuracy constraints.

[0035] Specifically, a first consultation request reply sample partition is selected and extracted from multiple consultation request reply sample partitions. Check whether the amount of data in the first consultation request reply sample partition, that is, the first sample partition data amount, is less than the preset partition data amount. This preset partition data amount is set in advance based on business requirements, model training requirements and other factors. If it is found that the amount of data in the first sample partition is less than the preset partition data amount, it means that the number of samples in the current partition is insufficient to support effective model training. At this time, it is necessary to perform a sample retrieval with the same consultation business attribute for this partition. The sample retrieval with the same consultation business attribute refers to finding additional samples with the same or similar business attributes as the first consultation request reply sample partition from global data or other available resources. Through the retrieval, a first supplementary dataset of request reply samples with the same attributes is generated, and this dataset will be used to supplement and expand the first consultation request reply sample partition. The first supplementary dataset of request reply samples with the same attributes is used to expand the data of the first consultation request reply sample partition, that is, to merge the two datasets to generate the first expanded request reply sample partition. The first expanded request reply sample partition now contains more sample data, which can meet the data volume requirements of model training. The first expansion request reply sample partition is trained and tested using multiple feature machine learning algorithms. These algorithms can be based on different machine learning technologies such as decision trees, support vector machines, neural networks, etc. Through training and testing, a first business consultation reply branch is generated. This branch is specially trained for the business attributes of the first consultation request reply sample partition, so it can more effectively process consultation requests with the same or similar attributes. Add the trained first business consultation reply branch to the multiple business consultation reply branches previously constructed. In this way, when the system receives a consultation request that meets the processing attributes of the first business consultation reply branch, this branch can be called for reply analysis. The first business consultation reply branch contains multiple first attribute business consultation responders that meet the consultation reply accuracy constraints. These responders are trained based on machine learning algorithms and can ensure the accuracy and efficiency of the replies.

[0036] Furthermore, step seven of this application also includes: According to the human-computer linkage interface, receiving a reply feedback result of the f-th user; When the reply feedback result includes newly added consultation request information, the newly added consultation request information is further analyzed based on the multiple business consultation semantic analysis branch, the consultation sensitivity evaluation branch and the business consultation reply channel to generate a newly added consultation reply result; According to the human-computer interaction interface, the newly added consultation reply result is encrypted and sent to the fth user.

[0037] Specifically, the feedback result of the fth user on the previous consultation reply is received through the human-computer interaction interface. This feedback includes the user's satisfaction with the reply, the correctness evaluation, or contains the newly added consultation request information. Check the received feedback result to determine whether it contains the newly added consultation request information. The newly added consultation request can be a further question or supplementary information raised by the user based on the previous reply. Once the newly added consultation request is identified, the consultation reply analysis of these new requests will continue based on the previously constructed multi-business consultation semantic parsing branch, consultation sensitivity evaluation branch and business consultation reply channel. The new consultation request is semantically parsed using the multi-business consultation semantic parsing branch to extract key information and intent. Then, the sensitivity of the new request is evaluated through the consultation sensitivity evaluation branch to ensure that the processing process meets security and compliance requirements. Finally, the matching business consultation reply branch in the business consultation reply channel analyzes the new request and generates a new consultation reply result. Before sending the new consultation reply result to the fth user, the reply result will be encrypted to ensure the security of data transmission. Encryption can use symmetric encryption, asymmetric encryption or hybrid encryption technology, depending on the security requirements and configuration of the system. After encryption is completed, the encrypted new consultation reply result is sent to the fth user through the human-computer linkage interface. After receiving the encrypted reply, the user can use the corresponding decryption key to decrypt and view it.

[0038] In summary, the intelligent business consulting method for mobile banking provided by this application has the following technical effects: By receiving the concurrent consultation request distribution according to the human-machine interaction interface of the target mobile bank, wherein the concurrent consultation request distribution includes a plurality of user consultation request information corresponding to a plurality of users; performing load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint to obtain a concurrent request response instruction; based on the concurrent request response instruction, performing business consultation priority analysis on the plurality of users to obtain a user business consultation priority constraint sequence, and traversing and extracting the concurrent consultation request distribution according to the user business consultation priority constraint sequence to generate the fth user consultation request information corresponding to the fth user, wherein f is a positive integer, f belongs to F, and F is the total number of users of the plurality of users; activating the multi-business consultation semantic analysis branch to perform semantic recognition on the fth user consultation request information to generate the fth consultation semantic feature data; according to the consultation The inquiry sensitivity evaluation branch performs deep sensitive analysis on the fth inquiry semantic feature data to generate an fth inquiry activation signal; based on the fth inquiry activation signal, activates the business inquiry reply channel to perform inquiry reply analysis on the fth inquiry semantic feature data to generate an fth inquiry reply result; according to the human-machine linkage interface, encrypts the fth inquiry reply result and sends it to the fth user, which effectively solves the problems that the existing technology is difficult to efficiently process massive data, the priority setting is rigid and cannot flexibly respond to changing needs, the understanding of user semantic intentions is not accurate enough to cause deviation or inaccuracy in reply content, and there is a lack of necessary protection mechanisms to ensure data security when processing sensitive information. By comprehensively optimizing concurrent processing, user priority, semantic understanding, sensitive information protection and data transmission security, the efficiency, security and user experience of mobile banking services are significantly improved.

[0039] Embodiment 2 Based on the same inventive concept as the intelligent business consulting method for mobile banking in the aforementioned embodiment, the present application also provides an intelligent business consulting device for mobile banking, see the attached Figure 2 , the device comprises: A request information acquisition module 11, the request information acquisition module 11 is used to receive a concurrent consultation request distribution according to a human-machine interaction interface of a target mobile bank, wherein the concurrent consultation request distribution includes a plurality of user consultation request information corresponding to a plurality of users; A response instruction acquisition module 12, the response instruction acquisition module 12 is used to perform load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint, and obtain a concurrent request response instruction; A constraint sequence acquisition module 13, the constraint sequence acquisition module 13 is used to perform business consultation priority analysis on the multiple users based on the concurrent request response instruction, obtain a user business consultation priority constraint sequence, and traverse and extract the concurrent consultation request distribution according to the user business consultation priority constraint sequence to generate the fth user consultation request information corresponding to the fth user, wherein f is a positive integer, f belongs to F, and F is the total number of users of the multiple users; A semantic recognition module 14, the semantic recognition module 14 is used to activate the multi-business consultation semantic analysis branch to perform semantic recognition on the f-th user consultation request information, and generate the f-th consultation semantic feature data; An activation signal generating module 15, the activation signal generating module 15 is used to perform a deep sensitive analysis on the fth consultation semantic feature data according to the consultation sensitive evaluation branch to generate an fth consultation activation signal; A reply result generating module 16, wherein the reply result generating module 16 is used to activate the business consultation reply channel to perform consultation reply analysis on the fth consultation semantic feature data according to the fth consultation activation signal, and generate the fth consultation reply result; The result sending module 17 is used to encrypt and send the fth consultation reply result to the fth user according to the human-computer linkage interface.

[0040] Furthermore, the response instruction acquisition module 12 in the device is also used for: Generating concurrent request capacity according to the concurrent consultation request distribution; Calculating a concurrent consultation load factor based on the concurrent request constraint capacity and the concurrent request capacity; Determining whether the concurrent consultation load factor satisfies the concurrent consultation load constraint; If the concurrent consultation load factor does not satisfy the concurrent consultation load constraint, generating a concurrent request warning instruction; If the concurrent consultation load factor satisfies the concurrent consultation load constraint, the concurrent request response instruction is obtained.

[0041] Furthermore, the semantic recognition module 14 in the device is also used for: The multiple business consultation semantic analysis branch includes a request voice noise reduction branch, a request voice recognition branch and a request semantic analysis branch; When the f-th user consultation request information is a voice signal, performing noise cleaning on the f-th user consultation request information according to the request voice noise reduction branch to generate f-th cleaned voice request information; Inputting the fth cleaning voice request information into the request voice recognition branch to obtain the fth voice recognition result; The fth speech recognition result is subjected to deep semantic analysis according to the request semantic analysis branch to obtain the fth consultation semantic feature data, wherein the request semantic analysis branch includes a consultation word semantic analysis layer, a consultation explicit semantic analysis layer and a consultation implicit semantic analysis layer.

[0042] Furthermore, the activation signal generating module 15 in the device is also used for: The consultation sensitivity evaluation branch includes a sensitive business consultation library, a consultation sensitivity identification branch and a consultation sensitivity judgment branch; According to the consultation sensitivity identification branch, correlation analysis is performed on the fth consultation semantic feature data and each sensitive business consultation feature data in the sensitive business consultation library to generate a plurality of feature sensitive consultation coefficients; Input the plurality of feature-sensitive consulting coefficients into the consulting sensitivity judgment branch, and determine whether the plurality of feature-sensitive consulting coefficients are less than a sensitive consulting threshold in the consulting sensitivity judgment branch; If any one of the plurality of characteristic sensitive consultation coefficients is greater than / equal to the sensitive consultation threshold, generating an f-th consultation taboo signal; If the plurality of characteristic sensitive consultation coefficients are all smaller than the sensitive consultation threshold, the fth consultation activation signal is obtained.

[0043] Furthermore, the reply result generating module 16 in the device is also used for: Extracting the fth consulting service attribute according to the fth consulting semantic feature data; Based on the fth consultation activation signal, activating the matching business consultation reply branch corresponding to the fth consultation business attribute in the business consultation reply channel; Perform consultation reply analysis on the fth consultation semantic feature data according to each business consultation responder in the matching business consultation reply branch to obtain multiple initial consultation reply results; Data fusion is performed based on the multiple initial consultation response results to obtain the fth consultation response result.

[0044] Furthermore, the device further comprises a channel construction module, wherein the channel construction module is used for: Performing a global consultation request reply sample search according to the target mobile bank to obtain a business consultation request reply record set, and performing data cleaning on the business consultation request reply record set to generate a business consultation request reply data set; Classifying consulting business attributes according to the business consulting request reply data set to obtain multiple consulting request reply sample partitions corresponding to multiple sample consulting business attributes; Based on the neural network, a business consultation reply channel architecture is constructed, wherein the business consultation reply channel architecture includes an input layer, a hidden layer and an output layer; Based on the multiple consultation request response sample partitions, training multiple business consultation response branches; The multiple business consultation reply branches are embedded into the hidden layer to generate a consultation reply analysis layer, and the input layer and the output layer are combined to generate the business consultation reply channel.

[0045] Furthermore, the device further comprises a supplementary data set generating module, wherein the supplementary data set generating module is used to: Extracting a first consultation request reply sample partition according to the plurality of consultation request reply sample partitions; Determine whether the first sample partition data volume of the first consultation request reply sample partition is less than the preset partition data volume; If the data volume of the first sample partition is less than the data volume of the preset partition, searching for samples with the same consulting business attributes for the first consultation request reply sample partition to generate a first supplementary data set of samples with the same attributes; Using the first same-attribute request reply sample supplementary data set to expand the first consultation request reply sample partition, to generate a first expanded request reply sample partition; According to multiple feature machine learning algorithms, the first expansion request reply sample partition is trained and tested respectively to generate a first business consultation reply branch, and the first business consultation reply branch is added to the multiple business consultation reply branches, wherein the first business consultation reply branch includes multiple first attribute business consultation responders that meet the consultation reply accuracy constraints.

[0046] Furthermore, the device further includes a newly added reply result acquisition module, and the newly added reply result acquisition module is used to: According to the human-computer linkage interface, receiving a reply feedback result of the f-th user; When the reply feedback result includes newly added consultation request information, the newly added consultation request information is further analyzed based on the multiple business consultation semantic analysis branch, the consultation sensitivity evaluation branch and the business consultation reply channel to generate a newly added consultation reply result; According to the human-computer interaction interface, the newly added consultation reply result is encrypted and sent to the fth user.

[0047] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1The intelligent business consulting method and specific examples for mobile banking in the first embodiment are also applicable to the intelligent business consulting device for mobile banking in this embodiment. Through the detailed description of the intelligent business consulting method for mobile banking, those skilled in the art can clearly understand the intelligent business consulting device for mobile banking in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0048] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0049] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. An intelligent business consulting method for mobile banking, characterized in that: The method comprises: receiving, according to the human-machine interaction interface of the target mobile bank, a concurrent consultation request distribution, wherein the concurrent consultation request distribution includes a plurality of user consultation request information corresponding to a plurality of users; Performing load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint to obtain concurrent request response instructions; Based on the concurrent request response instruction, business consultation priority analysis is performed on the multiple users to obtain a user business consultation priority constraint sequence, and the concurrent consultation request distribution is traversed and extracted according to the user business consultation priority constraint sequence to generate the fth user consultation request information corresponding to the fth user, where f is a positive integer, f belongs to F, and F is the total number of users of the multiple users; Activate the multi-business consultation semantic analysis branch to perform semantic recognition on the f-th user consultation request information to generate the f-th consultation semantic feature data; Performing a deep sensitive analysis on the fth consultation semantic feature data according to the consultation sensitive evaluation branch to generate an fth consultation activation signal; According to the fth consultation activation signal, activating the business consultation reply channel to perform consultation reply analysis on the fth consultation semantic feature data to generate the fth consultation reply result; The fth consultation reply result is encrypted and sent to the fth user according to the human-computer linkage interface.

2. The method according to claim 1, characterized in that Performing load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint to obtain a concurrent request response instruction includes: Generating concurrent request capacity according to the concurrent consultation request distribution; Calculating a concurrent consultation load factor based on the concurrent request constraint capacity and the concurrent request capacity; Determining whether the concurrent consultation load factor satisfies the concurrent consultation load constraint; If the concurrent consultation load factor does not satisfy the concurrent consultation load constraint, generating a concurrent request warning instruction; If the concurrent consultation load factor satisfies the concurrent consultation load constraint, the concurrent request response instruction is obtained.

3. The method according to claim 1, characterized in that Activate the multi-business consultation semantic analysis branch to perform semantic recognition on the f-th user consultation request information, and generate the f-th consultation semantic feature data, including: The multiple business consultation semantic analysis branch includes a request voice noise reduction branch, a request voice recognition branch and a request semantic analysis branch; When the f-th user consultation request information is a voice signal, performing noise cleaning on the f-th user consultation request information according to the request voice noise reduction branch to generate f-th cleaned voice request information; Inputting the fth cleaning voice request information into the request voice recognition branch to obtain the fth voice recognition result; The fth speech recognition result is subjected to deep semantic analysis according to the request semantic analysis branch to obtain the fth consultation semantic feature data, wherein the request semantic analysis branch includes a consultation word semantic analysis layer, a consultation explicit semantic analysis layer and a consultation implicit semantic analysis layer.

4. The method according to claim 1, characterized in that Performing a deep sensitive analysis on the fth consultation semantic feature data according to the consultation sensitive evaluation branch to generate the fth consultation activation signal includes: The consultation sensitivity evaluation branch includes a sensitive business consultation library, a consultation sensitivity identification branch and a consultation sensitivity judgment branch; According to the consultation sensitivity identification branch, correlation analysis is performed on the fth consultation semantic feature data and each sensitive business consultation feature data in the sensitive business consultation library to generate a plurality of feature sensitive consultation coefficients; Input the plurality of feature-sensitive consulting coefficients into the consulting sensitivity judgment branch, and determine whether the plurality of feature-sensitive consulting coefficients are less than a sensitive consulting threshold in the consulting sensitivity judgment branch; If any one of the plurality of characteristic sensitive consultation coefficients is greater than / equal to the sensitive consultation threshold, generating an f-th consultation taboo signal; If the plurality of characteristic sensitive consultation coefficients are all smaller than the sensitive consultation threshold, the fth consultation activation signal is obtained.

5. The method according to claim 1, characterized in that According to the fth consultation activation signal, the business consultation reply channel is activated to perform consultation reply analysis on the fth consultation semantic feature data to generate the fth consultation reply result, including: Extracting the fth consulting service attribute according to the fth consulting semantic feature data; Based on the fth consultation activation signal, activating the matching business consultation reply branch corresponding to the fth consultation business attribute in the business consultation reply channel; Perform consultation reply analysis on the fth consultation semantic feature data according to each business consultation responder in the matching business consultation reply branch to obtain multiple initial consultation reply results; Data fusion is performed based on the multiple initial consultation response results to obtain the fth consultation response result.

6. The method according to claim 1, characterized in that The steps of constructing the business consultation reply channel include: Performing a global consultation request reply sample search according to the target mobile bank to obtain a business consultation request reply record set, and performing data cleaning on the business consultation request reply record set to generate a business consultation request reply data set; Classifying consulting business attributes according to the business consulting request reply data set to obtain multiple consulting request reply sample partitions corresponding to multiple sample consulting business attributes; Based on the neural network, a business consultation reply channel architecture is constructed, wherein the business consultation reply channel architecture includes an input layer, a hidden layer and an output layer; Based on the multiple consultation request response sample partitions, training multiple business consultation response branches; The multiple business consultation reply branches are embedded into the hidden layer to generate a consultation reply analysis layer, and the input layer and the output layer are combined to generate the business consultation reply channel.

7. The method according to claim 6, characterized in that Based on the plurality of consultation request reply sample partitions, a plurality of business consultation reply branches are trained, including: Extracting a first consultation request reply sample partition according to the plurality of consultation request reply sample partitions; Determine whether the first sample partition data volume of the first consultation request reply sample partition is less than the preset partition data volume; If the data volume of the first sample partition is less than the data volume of the preset partition, searching for samples with the same consulting business attributes for the first consultation request reply sample partition to generate a first supplementary data set of samples with the same attributes; Using the first same-attribute request reply sample supplementary data set to expand the first consultation request reply sample partition, to generate a first expanded request reply sample partition; According to multiple feature machine learning algorithms, the first expansion request reply sample partition is trained and tested respectively to generate a first business consultation reply branch, and the first business consultation reply branch is added to the multiple business consultation reply branches, wherein the first business consultation reply branch includes multiple first attribute business consultation responders that meet the consultation reply accuracy constraints.

8. The method according to claim 1, characterized in that The method further comprises: encrypting the fth consultation reply result and sending it to the fth user according to the human-machine linkage interface; According to the human-computer linkage interface, receiving a reply feedback result of the f-th user; When the reply feedback result includes newly added consultation request information, the newly added consultation request information is further analyzed based on the multiple business consultation semantic analysis branch, the consultation sensitivity evaluation branch and the business consultation reply channel to generate a newly added consultation reply result; According to the human-computer interaction interface, the newly added consultation reply result is encrypted and sent to the fth user.

9. An intelligent business consulting device for mobile banking, characterized in that: For implementing the steps of the method according to any one of claims 1 to 8, the device comprises: A request information acquisition module, the request information acquisition module is used to receive a concurrent consultation request distribution according to a human-machine interaction interface of a target mobile bank, wherein the concurrent consultation request distribution includes a plurality of user consultation request information corresponding to a plurality of users; A response instruction acquisition module, the response instruction acquisition module is used to perform load verification on the concurrent consultation request distribution according to the concurrent consultation load constraint, and obtain a concurrent request response instruction; A constraint sequence acquisition module, the constraint sequence acquisition module is used to perform business consultation priority analysis on the multiple users based on the concurrent request response instruction, obtain a user business consultation priority constraint sequence, and traverse and extract the concurrent consultation request distribution according to the user business consultation priority constraint sequence to generate the fth user consultation request information corresponding to the fth user, wherein f is a positive integer, f belongs to F, and F is the total number of users of the multiple users; A semantic recognition module, the semantic recognition module is used to activate the multi-business consultation semantic analysis branch to perform semantic recognition on the f-th user consultation request information to generate the f-th consultation semantic feature data; an activation signal generating module, the activation signal generating module being used to perform a deep sensitive analysis on the fth consultation semantic feature data according to the consultation sensitive evaluation branch to generate an fth consultation activation signal; A reply result generating module, the reply result generating module is used to activate the business consultation reply channel to perform consultation reply analysis on the fth consultation semantic feature data according to the fth consultation activation signal, and generate the fth consultation reply result; A result sending module is used to encrypt and send the fth consultation reply result to the fth user according to the human-computer linkage interface.

Images