Question processing method and device, storage medium and electronic equipment

Abstract

The invention discloses a problem processing method and device, a storage medium and electronic equipment. The method relates to the field of artificial intelligence, and comprises the following steps: under the condition that a target question of a user is received, performing retrieval in a target knowledge base based on the target question to obtain a retrieval result; under the condition that the retrieval result comprises at least one piece of reply information, inputting the at least one piece of reply information into a target neural network model, and determining first reply information for the target question through the target neural network model; under the condition that the retrieval result does not include the reply information, determining an artificial customer service corresponding to the user, transferring the artificial customer service to the user, and sending first prompt information to the artificial customer service; and under the condition that the target work order triggered by the manual customer service is detected, generating second prompt information according to the target work order, and sending the second prompt information to the first target object. The technical problem that the answering efficiency is low due to the fact that the user questions are answered by depending on manual customer service in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a problem processing method, device, storage medium and electronic device. Background Art

[0002] In the traditional public account customer service dialogue process, it is usually necessary to rely on manual access to respond to user inquiries. However, the traditional customer service model requires manual customer service to be on duty 24 hours a day, 7 days a week to ensure a positive response to users. This model that relies on manpower is not only costly, but also easily causes users to wait too long. Especially during peak hours when customer service staff is limited, manual customer service will be too busy to answer questions too slowly, resulting in low efficiency in answering user questions.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present invention provide a problem handling method, device, storage medium and electronic device to at least solve the technical problem in the related art of relying on manual customer service to answer user questions, resulting in low answering efficiency.

[0005] According to one aspect of an embodiment of the present invention, a problem handling method is provided, including: when a target question of a user is received, searching in a target knowledge base based on the target question to obtain a search result, wherein the target knowledge base includes multiple question-answer pairs, and the question-answer pairs include questions and reply information corresponding to the questions; when at least one reply information is included in the search result, inputting the at least one reply information into a target neural network model, and determining first reply information for the target question through the target neural network model; when the search result does not include reply information, determining the manual customer service corresponding to the user, transferring the user to the manual customer service, and sending a first prompt information to the manual customer service, wherein the first prompt information is used to prompt the manual customer service to answer the target question; when a target work order triggered by the manual customer service is detected, generating a second prompt information according to the target work order, and sending the second prompt information to a first target object, wherein the target work order includes a target question, and the second prompt information is used to prompt the first target object to answer the target question.

[0006] Furthermore, the target work order also includes information about the object to receive the target work order, wherein the problem handling method also includes: determining the first target object according to the information about the object to receive the target work order in the target work order; and sending a second prompt message to the first target object.

[0007] Furthermore, the problem handling method also includes: after sending the second prompt information to the first target object, receiving the work order processing result fed back by the first target object, and sending the work order processing result to the manual customer service; when receiving the knowledge base update request sent by the manual customer service, extracting the second reply information for the target problem from the work order processing result, and updating the target knowledge base based on the target problem and the second reply information.

[0008] Furthermore, the problem handling method also includes: after sending the second prompt message to the first target object, in the case of receiving the work order transfer request sent by the first target object, determining the second target object of the target work order to be transferred from the work order transfer request, generating a third prompt message according to the target work order, and sending the third prompt message to the second target object, wherein the third prompt message is used to prompt the second target object to answer the target question; in the case of receiving the work order cancellation request sent by the manual customer service, generating a fourth prompt message according to the work order information of the target work order, and sending the fourth prompt message to the first target object, wherein the fourth prompt message is used to prompt the first target object to cancel the processing of the target work order; determining the time point of sending the second prompt message as the target time point, and calculating the first time length between the current time point and the target time point, and when the first time length meets the preset time length condition, sending the second prompt message to the first target object again.

[0009] Furthermore, the problem handling method also includes: before searching in the target knowledge base based on the target question and obtaining the search results, upon receiving the target information sent by the user, sending first information to the user, wherein the target information is information that does not include the question, and the first information is used to prompt the user to select the user's subject type; receiving the subject type fed back by the user based on the first information, and determining multiple question types corresponding to the subject type, generating second information based on the multiple question types, and sending the second information to the user, wherein the second information is used to prompt the user to select the question type; receiving the question type fed back by the user based on the second information, and determining multiple candidate questions corresponding to the question type, generating third information based on the multiple candidate questions, and sending the third information to the user, wherein the third information is used to prompt the user to select a candidate question; receiving the question fed back by the user based on the third information, and determining the question as the target question.

[0010] Furthermore, the problem processing method also includes: when the target question does not belong to multiple candidate questions, searching in the target knowledge base based on the target question to obtain a search result.

[0011] Furthermore, the question processing method also includes: after determining the question as the target question, if the target question belongs to multiple candidate questions, determining the preset reply information corresponding to the candidate question that is the same as the target question as the reply information for the target question.

[0012] According to another aspect of an embodiment of the present invention, a problem handling device is also provided, including: a retrieval module, which is used to search in a target knowledge base based on the target question when a target question of a user is received, and obtain a retrieval result, wherein the target knowledge base includes multiple question-answer pairs, and the question-answer pairs include questions and answer information corresponding to the questions; a first determination module, which is used to input at least one answer information into a target neural network model when at least one answer information is included in the retrieval result, and determine the first answer information for the target question through the target neural network model; a second determination module, which is used to determine the manual customer service corresponding to the user when the retrieval result does not include the answer information, transfer the manual customer service to the user, and send a first prompt information to the manual customer service, wherein the first prompt information is used to prompt the manual customer service to answer the target question; a first sending module, which is used to generate a second prompt information according to the target work order when a target work order triggered by the manual customer service is detected, and send the second prompt information to the first target object, wherein the target work order includes the target question, and the second prompt information is used to prompt the first target object to answer the target question.

[0013] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided, in which a computer program is stored. The computer program is configured to execute the above-mentioned problem handling method when running.

[0014] According to another aspect of an embodiment of the present invention, an electronic device is also provided, which includes one or more processors; a memory for storing one or more programs, so that when the one or more programs are executed by the one or more processors, the one or more processors are implemented to run the programs, wherein the programs are configured to execute the above-mentioned problem handling method when running.

[0015] In an embodiment of the present invention, a method of combining intelligent customer service and manual customer service to answer user questions is adopted. When receiving the user's target question, a search is performed in the target knowledge base based on the target question to obtain a search result. Then, when the search result includes at least one reply information, the at least one reply information is input into the target neural network model, and the first reply information for the target question is determined by the target neural network model. Then, when the search result does not include the reply information, the manual customer service corresponding to the user is determined, the user is transferred to the manual customer service, and a first prompt information is sent to the manual customer service. When a target work order triggered by the manual customer service is detected, a second prompt information is generated according to the target work order, and the second prompt information is sent to the first target object. The target knowledge base includes multiple question-answer pairs, the question-answer pairs include questions and reply information corresponding to the questions, the target work order includes the target question, and the second prompt information is used to prompt the first target object to answer the target question.

[0016] In the above process, when the target question of the user is received, the system automatically searches and generates reply information, so as to provide services to the user by means of intelligent customer service. When the reply information is not included in the search results, the manual customer service corresponding to the user is determined and the user is transferred to the manual customer service. When the intelligent customer service cannot provide accurate answers to the user, the manual customer service can be effectively connected to provide services to the user, thereby placing the intelligent customer service consultation service before the manual customer service, ensuring that the intelligent customer service is given priority and the manual customer service assists in answering. Since the intelligent customer service can reply to the user's questions relatively quickly without waiting in line, it can effectively improve the efficiency of answering the user's questions, shorten the user's waiting time, and reduce labor costs. In addition, when the target work order triggered by the manual customer service is detected, the second prompt information is sent to the first target object to prompt the first target object to answer the target question. When the manual customer service cannot provide accurate answers to the user, the system can quickly transfer the target question to other professionals, thereby further improving the efficiency and reliability of the answer.

[0017] It can be seen that the solution provided in this application achieves the purpose of combining intelligent customer service and manual customer service to answer user questions, thereby achieving the technical effect of improving answering efficiency, and further solving the technical problem of relying on manual customer service to answer user questions in related technologies, resulting in low answering efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0019] Figure 1is a schematic diagram of an optional problem handling method according to an embodiment of the present invention;

[0020] Figure 2 is a schematic diagram of an optional target processing system according to an embodiment of the present invention;

[0021] Figure 3 This is a flow chart of an optional problem solving method according to an embodiment of the present invention. Figure 1 ;

[0022] Figure 4 is a schematic diagram of an optional processing target work order according to an embodiment of the present invention Figure 1 ;

[0023] Figure 5 is a schematic diagram of an optional processing target work order according to an embodiment of the present invention Figure 2 ;

[0024] Figure 6 is a schematic diagram of an optional processing target work order according to an embodiment of the present invention Figure 3 ;

[0025] Figure 7 is a schematic diagram of an optional reminder of a first target object according to an embodiment of the present invention;

[0026] Figure 8 is a schematic diagram of an optional target problem determination according to an embodiment of the present invention;

[0027] Fig. 9 This is a flow chart of an optional problem solving method according to an embodiment of the present invention. Figure 2 ;

[0028] Fig.10 is a schematic diagram of an optional update target knowledge base according to an embodiment of the present invention;

[0029] Fig.11 is a schematic diagram of an optional problem handling device according to an embodiment of the present invention;

[0030] Fig.12 is a schematic diagram of an optional electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards in the relevant regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0034] Example 1

[0035] According to an embodiment of the present invention, an embodiment of a problem handling method is 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 a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0036] Figure 1 is a schematic diagram of an optional problem solving method according to an embodiment of the present invention, such as Figure 1 As shown, the method comprises the following steps:

[0037] Step S101, upon receiving a target question from a user, a search is performed in a target knowledge base based on the target question to obtain a search result, wherein the target knowledge base includes a plurality of question-answer pairs, and the question-answer pairs include questions and answer information corresponding to the questions.

[0038] Optionally, electronic devices, application systems, servers and other devices may be used as the execution subject of the present application. In this embodiment, the target processing system is used as the execution subject to execute the above-mentioned problem handling method.

[0039] Optional, Figure 2 is a schematic diagram of an optional target processing system according to an embodiment of the present invention. Figure 2 As shown, the target processing system includes at least an intelligent customer service subsystem, a work order management subsystem, and a customer service dialogue platform. Users raise target questions through the customer service dialogue platform, and the target processing system can answer the target questions through the intelligent customer service in the intelligent customer service subsystem on the customer service dialogue platform, or transfer to manual customer service, and have the manual customer service answer the target questions on the customer service dialogue platform. Figure 2 As shown, the intelligent customer service can determine the first reply information for the target question based on the knowledge in the target knowledge base. If it cannot be answered, it can be transferred to the manual customer service for answer. If the manual customer service cannot solve the target question during the process of answering the target question, it can send a work order to a professional, and the professional will help answer the target question. In addition, the manual customer service can also maintain the knowledge in the target knowledge base.

[0040] In some embodiments, the target processing system may register the customer service dialogue platform to the relevant dialogue development platform, successfully register the intelligent customer service, and configure relevant functions for the intelligent customer service to provide question answering functions for users on the relevant dialogue development platform. In addition, the target processing system may also establish an offline knowledge base based on at least part of the knowledge in the target knowledge base, and import the offline knowledge base into the customer service dialogue platform.

[0041] Optionally, the content of the user's target question may vary depending on the actual application scenario. Optionally, the problem processing method provided in this embodiment may be applied to scenarios such as power trading, e-commerce, and financial product sales. In the case of a power trading scenario, the target question may be "What is the website address of the power trading center trading platform?" In this embodiment, the content of the user's target question is not specifically limited.

[0042] Optionally, when receiving the user's target question, the intelligent customer service subsystem in the target processing system can search the target knowledge base based on the target question to obtain a search result, wherein the target knowledge base includes multiple question-answer pairs, and the question-answer pairs include questions and answer information corresponding to the questions. For example, the intelligent customer service subsystem can perform vector conversion on the target question to obtain a question vector, and obtain the question-answer pair vector corresponding to each question-answer pair in the target knowledge base, thereby screening out question-answer pairs with a similarity greater than a preset similarity from multiple question-answer pairs based on the similarity between the question vector and the question-answer pair vector, and determining the search result based on the answer information in the question-answer pairs with a similarity greater than the preset similarity. Optionally, when there is no question-answer pair with a similarity greater than a preset similarity among the multiple question-answer pairs, the answer information is not included in the search result.

[0043] Step S102, when the search results include at least one reply information, input the at least one reply information into the target neural network model, and determine the first reply information for the target question through the target neural network model.

[0044] Optionally, when at least one reply information is included in the search results, it is determined that there is reply information in the historical replies that can solve the target problem. In this case, the intelligent customer service subsystem may input the at least one reply information and the target problem into the target neural network model, thereby determining the first reply information for the target problem through the target neural network model. For example, the target neural network model is implemented based on natural language processing (Natural Language Processing, NLP) technology, and the target neural network model may be a large language model. The intelligent customer service subsystem may input at least one reply information, the target problem, and an instruction statement into the target neural network model, wherein the instruction statement is used to guide the target neural network model to generate the first reply information for the target problem according to the at least one reply information.

[0045] Optionally, based on steps S101-S102, the intelligent customer service subsystem can provide standardized responses to ensure the consistency and accuracy of the responses. This is very effective in handling common problems and avoids the repetitive work of manual customer service answering repeated questions. In addition, by using neural network models and natural language processing technology, the intelligent customer service subsystem can quickly learn and adapt to new situations and problems, and continuously improve its service level. When answering user questions through intelligent customer service, the intelligent customer service subsystem has the advantages of high efficiency, low cost, stable service quality, simple operation, time saving and labor saving compared to traditional customer service.

[0046] Step S103, when the search results do not include reply information, determine the manual customer service corresponding to the user, transfer the user to the manual customer service, and send a first prompt message to the manual customer service, wherein the first prompt message is used to prompt the manual customer service to answer the target question.

[0047] Optionally, when the search results do not include reply information, it is determined that there is no reply information in the historical replies that can solve the target problem. In this case, the intelligent customer service subsystem in the target processing system can determine the human customer service corresponding to the user. For example, a human customer service is randomly determined from multiple human customer services as the human customer service corresponding to the user. For another example, the human customer service that currently serves the least number of users is found from multiple human customer services, and the human customer service is determined as the human customer service corresponding to the user.

[0048] Optionally, after determining the manual customer service corresponding to the user, the intelligent customer service subsystem can transfer the user to the manual customer service and send a first prompt message to the manual customer service, wherein the first prompt message is used to prompt the manual customer service to answer the target question. The first prompt message may include the user information of the user, for example, the first prompt message may be "Please provide manual service for user A".

[0049] Step S104, when a target work order triggered by manual customer service is detected, second prompt information is generated according to the target work order, and the second prompt information is sent to the first target object, wherein the target work order includes a target question, and the second prompt information is used to prompt the first target object to answer the target question.

[0050] Optionally, in the process of manual customer service providing customer service to users, the manual customer service may be unable to solve the user's target problem. In this case, the manual customer service can initiate a target work order through the work order management subsystem. Optionally, the target work order includes at least the target problem and the customer service information of the manual customer service, and the work order management subsystem can randomly determine a target object from multiple preset target objects as the first target object. Optionally, in the case where the target work order includes information about the object to receive the target work order, the work order management subsystem can determine the first target object based on the information about the object to receive the target work order in the target work order, and then send a second prompt message to it. For example, the second prompt message can be "Please answer the question recorded in the target work order (i.e., the target problem)".

[0051] In some embodiments, the first target audience is professionals in the technical field to which the target problem belongs.

[0052] For example, Figure 3 This is a flow chart of an optional problem solving method according to an embodiment of the present invention. Figure 1 ,like Figure 3As shown, after the user raises the target question, the intelligent customer service in the intelligent customer service subsystem will first access it, and then search the target knowledge base according to the target question, and judge whether the target question can be solved according to the search results. If the problem can be solved, the target question will be intelligently replied. If the target problem cannot be solved, it will be transferred to the manual customer service, and the manual customer service will answer it. If the manual customer service cannot answer it either, the manual customer service can initiate a target work order, and the work order management subsystem will send the target work order to the professional (i.e., the first target object) for answering. If the manual customer service can answer it, the manual customer service will answer the question directly. After that, the manual customer service can also maintain the knowledge base information based on this answer. After the user's inquiry is completed, the target processing system can initiate a user evaluation request to the user. This completes the customer service service.

[0053] Based on the scheme defined in the above steps S101 to S104, it can be known that in an embodiment of the present invention, a method of combining intelligent customer service and manual customer service to answer user questions is adopted. When receiving the user's target question, a search is performed in the target knowledge base based on the target question to obtain a search result. Then, when the search result includes at least one reply information, the at least one reply information is input into the target neural network model, and the first reply information for the target question is determined by the target neural network model. Then, when the search result does not include the reply information, the manual customer service corresponding to the user is determined, the user is transferred to the manual customer service, and a first prompt information is sent to the manual customer service. Therefore, when a target work order triggered by the manual customer service is detected, a second prompt information is generated according to the target work order, and the second prompt information is sent to the first target object. Among them, the target knowledge base includes multiple question-answer pairs, the question-answer pairs include questions and reply information corresponding to the questions, the target work order includes the target question, and the second prompt information is used to prompt the first target object to answer the target question.

[0054] It is easy to notice that in the above process, by automatically searching and generating reply information by the system when receiving the user's target question, it is realized to provide services to users in the form of intelligent customer service. By determining the manual customer service corresponding to the user when the reply information is not included in the search results, and transferring the user to the manual customer service, it is realized that when the intelligent customer service cannot provide accurate answers to the user, the manual customer service can be effectively connected to provide services to the user, thereby realizing that the intelligent customer service acceptance consultation service is placed before the manual customer service, ensuring that the intelligent customer service is accepted first, and the manual customer service assists in answering. Since the intelligent customer service can reply to the user's questions relatively quickly without waiting in line, it can effectively improve the efficiency of answering user questions, shorten the user's waiting time, and reduce labor costs. In addition, by sending a second prompt message to the first target object to prompt the first target object to answer the target question when the manual customer service is detected to trigger the target work order, it is realized that when the manual customer service cannot provide an accurate answer to the user, the system can quickly transfer the target question to other professionals, thereby further improving the efficiency and reliability of the answer.

[0055] It can be seen that the solution provided in this application achieves the purpose of combining intelligent customer service and manual customer service to answer user questions, thereby achieving the technical effect of improving answering efficiency, and further solving the technical problem of relying on manual customer service to answer user questions in related technologies, resulting in low answering efficiency.

[0056] In an optional embodiment, the target work order also includes information about the object to receive the target work order, wherein, in the process of sending the second prompt information to the first target object, the target processing system can determine the first target object based on the information about the object to receive the target work order in the target work order, thereby sending the second prompt information to the first target object.

[0057] Optionally, the information of the object to receive the target work order includes the object name and the department to which the object belongs. The work order management subsystem in the target processing system can determine the first target object according to the object name and the department to which the object belongs, and then send the second prompt information to the first target object.

[0058] Optional, Figure 4 is a schematic diagram of an optional processing target work order according to an embodiment of the present invention Figure 1 ,like Figure 4 As shown, the manual customer service can click Add a Work Order in the Work Order Management Subsystem, and then fill in the target problem, fill in the customer service information of the manual customer service (such as customer service ID, customer service contact information, etc.) in the Work Order Management Subsystem, and select the department and name of the person to receive the target work order.

[0059] Optionally, the target work order generation method is divided into offline work order generation and online work order generation. Figure 4 As shown, after the manual customer service has determined the above-mentioned target question, customer service information, department and name, if the manual customer service selects the offline work order generation method, the front end of the work order management subsystem will download the corresponding file, which includes the information determined by the manual customer service, thereby determining the file as the target work order. Afterwards, the customer service staff can directly forward the generated target work order to the first target object through the work order management subsystem, or the customer service staff can perform secondary editing on the target work order and forward it to the first target object through the work order management subsystem. After the first target object obtains the corresponding work order, it can fill in the offline work order. After filling in, the work order processing result (such as the updated target work order) is obtained, and the work order processing result is uploaded to the work order management subsystem. At the same time, the work order status is changed to answered. As shown Figure 4 As shown, if the generation method selected by the manual customer service is the online work order generation method, the work order management subsystem can automatically generate the target work order in response to the user's instructions, and push the target work order to the first target object, and then the first target object fills it out online in the work order management subsystem, and clicks submit in the work order management subsystem after filling it out, so that the work order management subsystem obtains the work order processing result.

[0060] It should be noted that, through the above process, the first target object is accurately determined, thereby further improving the reliability of answering questions.

[0061] In an optional embodiment, after sending the second prompt information to the first target object, the target processing system may receive the work order processing result fed back by the first target object, and send the work order processing result to the manual customer service, and upon receiving the knowledge base update request sent by the manual customer service, extract the second reply information for the target question from the work order processing result, and update the target knowledge base based on the target question and the second reply information.

[0062] In some embodiments, the work order processing result may include second reply information for the target problem determined by the first target object, and the work order processing result may be a work order obtained after the first target object updates the target work order.

[0063] Optionally, after obtaining the work order processing result, the work order management subsystem can send the work order processing result to the manual customer service, so that the second reply information in the work order processing result can be fed back to the user through the manual customer service. If the customer service can quickly locate the user, he can reply in private chat. If the user cannot be located, the user notification function of the target processing system can be used to notify.

[0064] Optional, such as Figure 4As shown, after the first target object submits the work order processing result to the work order management subsystem, it can choose whether to notify the manual customer service in the work order management subsystem. If it is chosen to notify the manual customer service, the work order management subsystem can send notification information and the work order processing result to the manual customer service, wherein the notification information is used to notify the manual customer service that the work order processing result has been determined. If it is chosen not to notify the manual customer service, if it is not chosen to notify the manual customer service, the work order management subsystem can only send the work order processing result to the manual customer service.

[0065] Optional, such as Figure 4 As shown, after receiving the work order processing result, the customer service staff (i.e., manual customer service) can select an option to update the target knowledge base. In the case where the customer service staff chooses to update the target knowledge base, the customer service staff can send a request to update the knowledge base to the work order management subsystem, and then the work order management subsystem extracts the second reply information for the target question from the work order processing result, and updates the target knowledge base based on the target question and the second reply information, for example, a question-answer pair is formed based on the target question and the second reply information, and then the question-answer pair is stored in the target knowledge base. When the customer service staff chooses not to update the target knowledge base, the work order management subsystem does not operate on the target knowledge base.

[0066] In some embodiments, the work order processing result may not include reply information, but include rejection information, wherein the rejection information indicates that the first target object refuses to answer the target question. For example, when the technical field of the first target object is different from the technical field of the target question, the work order processing result generated by the first target object may include rejection information.

[0067] Optionally, after receiving the work order processing result, the customer service staff can choose whether to notify the user according to the actual situation, for example, if the work order processing result includes the second reply information, the user is notified. The customer service staff can select the notification method in the target processing system, for example, sending the second reply information to the questioning user through the customer service dialogue platform, or sending the second reply information to the questioning user through PC push, SMS reminder, etc.

[0068] Optionally, the developer of the target processing system can develop a question answering page and add permission judgment. For example, only the first target object can enter the page to submit answers (i.e., the work order processing results). Optionally, the developer can also develop a customer service page to view the details of the problem, and also add permission judgment. Only the customer service staff corresponding to the target work order can click to view the details page and update the target knowledge base. For example, Figure 5 is a schematic diagram of an optional processing target work order according to an embodiment of the present invention Figure 2 ,like Figure 5As shown, the customer service staff (i.e., manual customer service) creates a target work order, and then sends the target work order to the answering staff (i.e., the first target object) through the target processing system. When the answering staff views the target work order, the target processing system verifies the answering staff's authority (for example, sends a verification SMS, verifies through a verification code). If the answering authority exists, the answering staff is allowed to generate the work order processing result. If the answering authority does not exist, the answering staff is prohibited from operating. Afterwards, the work order processing result is fed back to the customer service staff. If the work order processing result includes rebuttal information, the customer service staff may not feedback information to the user. If the work order processing result does not include rebuttal information, the second reply information in the work order processing result is sent to the user. In addition, the customer service staff can also decide whether to update the target knowledge base.

[0069] It should be noted that by sending the work order processing results to manual customer service, manual customer service can promptly answer users' target questions. By updating the target knowledge base upon receiving a request to update the knowledge base, timely supplementation of knowledge in the target knowledge base is achieved, thereby facilitating the improvement of the scope of answers of intelligent customer service.

[0070] In an optional embodiment, after sending the second prompt message to the first target object, the target processing system can, upon receiving the work order transfer request sent by the first target object, determine the second target object of the target work order to be transferred from the work order transfer request, generate a third prompt message based on the target work order, and send the third prompt message to the second target object, wherein the third prompt message is used to prompt the second target object to answer the target question; upon receiving the work order cancellation request sent by the manual customer service, generate a fourth prompt message based on the work order information of the target work order, and send the fourth prompt message to the first target object, wherein the fourth prompt message is used to prompt the first target object to cancel the processing of the target work order; determine the time point of sending the second prompt message as the target time point, calculate the first time length between the current time point and the target time point, and send the second prompt message to the first target object again when the first time length meets the preset time length condition.

[0071] Optionally, after sending the second prompt message to the first target object, the first target object and the manual customer service can operate the target work order again. The operation methods include but are not limited to transfer, cancel and urge. For example, Figure 6 is a schematic diagram of an optional processing target work order according to an embodiment of the present invention Figure 3 ,like Figure 6As shown, after adding a target work order, if an emergency occurs, such as the answering person (i.e., the first target object) is temporarily out of the office and the answering person is unable to answer the question, the work order needs to be flexibly transferred by reselecting personnel from other professional departments. In this case, the first target object can initiate a work order transfer request in the work order management subsystem. The work order transfer request includes the information of the second target object of the target work order to be transferred (such as: object identification, etc.). When the work order management subsystem receives the work order transfer request, it can generate a third prompt message based on the target work order and send it to the second target object (i.e., Figure 6 The third prompt message is sent to the assigned personnel in the target work order. For example, the third prompt message may be "Please answer the question recorded in the target work order (i.e., the target question) on behalf of the first target object." After receiving the third prompt message, the second target object processes the target work order indicated in the third prompt message and submits the work order processing result to the work order management subsystem.

[0072] Optional, such as Figure 6 As shown, after adding a target work order, if an emergency occurs, such as the user who asks the question tells the manual customer service that he already knows the answer to the question, the manual customer service needs to cancel the work order and prevent subsequent operations on the work order. In this case, the manual customer service can initiate a work order cancellation request in the work order management subsystem, and the work order cancellation request includes at least the work order information of the target work order (such as: work order identification, etc.), so that the work order management subsystem generates a fourth prompt message based on the work order information of the target work order when receiving the work order cancellation request sent by the manual customer service, and sends the fourth prompt message to the first target object. For example, the fourth prompt message can be "no need to process the work order identified as A".

[0073] Optional, such as Figure 6 As shown, after adding a target work order, a notification will be sent to the answerer for the first time to remind him to answer the question, and the length of time that the answerer has not answered the question will be calculated in real time (i.e., the first time length between the current time point and the target time point). If the answerer has not answered the question in half an hour, the notification will be sent again until the answerer has finished answering the question. In addition, you can click the reminder button to send a notification once you click it. In the above case, the preset time length condition can be that the first time length reaches 30 minutes or a multiple of 30 minutes. For example, Figure 7 is a schematic diagram of an optional reminder of a first target object according to an embodiment of the present invention, such as Figure 7As shown, the work order management subsystem in the target processing system can scan all work orders in real time to determine whether there are unprocessed work orders. If so, it is determined whether the time for the answering personnel (i.e., the first target object) to not answer the question exceeds half an hour. If so, the second prompt information is sent to the answering personnel again. If not, the second prompt information is not sent to the first target object again. Optionally, if there are no unprocessed work orders in the system, wait for the next scan.

[0074] It should be noted that, through the above process, adaptive processing of the target work order is achieved under different circumstances, thereby improving the flexibility of processing the target work order, and further improving the reliability of answering user questions.

[0075] In an optional embodiment, before searching the target knowledge base based on the target question and obtaining the search results, the target processing system may, upon receiving the target information sent by the user, send first information to the user, wherein the target information is information that does not include a question, and the first information is used to prompt the user to select the user's subject type; receive the subject type fed back by the user based on the first information, determine multiple question types corresponding to the subject type, generate second information based on the multiple question types, and send the second information to the user, wherein the second information is used to prompt the user to select the question type; receive the question type fed back by the user based on the second information, determine multiple candidate questions corresponding to the question type, generate third information based on the multiple candidate questions, and send the third information to the user, wherein the third information is used to prompt the user to select a candidate question; receive the question fed back by the user based on the third information, and determine the question as the target question.

[0076] In some embodiments, in the power trading scenario, the subject type may be a power user, a power sales company, a power generation company, or other subjects. There is a corresponding relationship between the question type and the subject type. For example, if the subject type is a power user, the question type may be power user registration, basic information change, power usage information change, digital certificate binding, etc.

[0077] In some embodiments, the target information is information that does not include a question. For example, the target information may be “Hello”, “I want to ask some questions”, etc.

[0078] For example, Figure 8 is a schematic diagram of an optional target problem determination according to an embodiment of the present invention. The intelligent customer service subsystem in the target processing system can adopt Figure 8 The process shown in realizes guiding users to inquire from "market subject-question type-question selection", thereby determining the target question. Figure 8As shown, upon receiving the target information sent by the user, the intelligent customer service subsystem can send the first information to the user. For example, the first information is "I am the intelligent robot Xiaowei. Which of the following market entities are you? 1. Electricity user; 2. Power sales company; 3. Power generation enterprise; 4. Other entities". After the user selects the market entity according to the instructions, the corresponding multiple question types are determined according to the entity type fed back by the user, for example, the popular top 5 types under the entity type are determined, and then the second information is generated to ask the user what type of question they need to ask. For example, if the entity type is an electricity user, the second information can be "Do you want to ask the following questions? You can also enter other questions to ask more questions. 1. Electricity user registration; 2. Basic information change; 3. Electricity usage information change; 4. Digital certificate binding; 5. Others". After the user selects the question type, if the question type is other, the intelligent customer service subsystem can re-determine multiple question types corresponding to the subject type and regenerate the second information. If the question type is not other, the intelligent customer service subsystem can generate the third information based on the popular top 5 questions (i.e. candidate questions) under this question type. For example, if the question type is an introduction to the power trading platform, the third information can be "Do you want to know the following questions? 1. How to log in to the power trading center platform; 2. Introduction to power trading training courses; 3. Website of the power trading center trading platform". The user can select questions according to actual needs. If not, click Other to get other recommended questions. The user can select according to needs. If none of them are suitable, the user can finally enter his own question to inquire. The intelligent customer service subsystem can receive questions from users based on the third information feedback and determine the question as the target question.

[0079] In some embodiments, when receiving the subject type fed back by the user based on the first information, the intelligent customer service subsystem can determine whether the user has also fed back the question type. If the user has not fed back the question type, the above-mentioned second information is generated. If the user has fed back the question type, and the question type is a question type stored in the intelligent customer service subsystem, multiple candidate questions corresponding to the question type are directly determined, and the third information is generated based on the multiple candidate questions. If the question type is not a question type stored in the intelligent customer service subsystem, multiple question types corresponding to the subject type are determined, and the second information is generated based on the multiple question types.

[0080] It should be noted that, through the above process, the target question can be accurately determined, thereby improving the accuracy of the answer.

[0081] In an optional embodiment, in the process of searching the target knowledge base based on the target question to obtain the search result, the target processing system may search the target knowledge base based on the target question to obtain the search result when the target question does not belong to multiple candidate questions.

[0082] Optionally, the target processing system has preset answer information corresponding to each candidate question. When the target question does not belong to multiple candidate questions, the target processing system determines that the answer information corresponding to the target question cannot be directly obtained. In this case, the target processing system can search the target knowledge base based on the target question to obtain the search results, and then determine the way to answer the target question according to the search results, such as generating through the target neural network model, or answering through manual customer service.

[0083] It should be noted that by searching to determine the answer information for the target question when there is no answer information corresponding to the target question in the system, the problem of directly transferring the user to manual customer service in this case, which increases the workload of manual customer service and affects the efficiency of answering questions, is avoided.

[0084] In an optional embodiment, after determining the question as the target question, the target processing system may, if the target question belongs to multiple candidate questions, determine the preset answer information corresponding to the candidate question that is the same as the target question as the answer information for the target question.

[0085] Optionally, the target processing system has preset answer information corresponding to each candidate question. When the target question belongs to multiple candidate questions, the target processing system determines that the answer information corresponding to the target question can be directly obtained. Therefore, the target processing system can determine the preset answer information corresponding to the candidate question that is the same as the target question as the answer information for the target question.

[0086] It should be noted that, through the above process, the efficiency of answering user questions can be further improved.

[0087] In an optional embodiment, Fig. 9 This is a flow chart of an optional problem solving method according to an embodiment of the present invention. Figure 2 ,like Fig. 9As shown, after the user inputs the target question, the target processing system determines whether the answer information can be found in the target knowledge base to solve the problem. If so, the first answer information for the target question is generated through the target neural network model and fed back to the user. After the user's problem is solved, the user is asked whether he has other questions. If so, it is determined whether the answer information can be found in the target knowledge base to solve the new problem. If not, an end statement is sent to end the conversation, and if the user does not initiate another inquiry after 10 minutes, the user is actively invited to evaluate. Optionally, if the answer information cannot be found in the target knowledge base to solve the problem, it is determined that the user has entered an unknown question once. In this case, the target processing system can guide the user to ask the question again (such as sending "Please re-enter your question" to the user), and determine whether the answer information can be found in the target knowledge base to solve the problem re-entered by the user. If yes, the answer information for the problem is generated through the target neural network model. If not, it is determined that the user has entered the unknown question twice. In this case, the user is transferred to the manual customer service, and the user is invited to evaluate the service after the manual conversation ends.

[0088] Optionally, after determining the first reply information for the target question, the target processing system may feed back the first reply information to the user and ask the user whether to finish asking the question. If the user finishes asking the question, the end phrase is sent, and if the user does not ask the question again after 10 minutes, the user is actively invited to make an evaluation. If the user sends a message again within 10 minutes, the user's question will continue to be answered.

[0089] Optionally, after getting an answer through manual inquiry, the session will be ended by the manual customer service, or it will be ended automatically after 30 minutes, after which the user will receive an invitation link to evaluate.

[0090] In some embodiments, Fig.10 is a schematic diagram of an optional update target knowledge base according to an embodiment of the present invention, such as Fig.10 As shown, the methods for updating the target knowledge base include a single update method and a batch update method. When the update method is a single update method, the manual customer service can send a request to update the knowledge base to the target processing system after receiving the work order processing result, so as to update the target knowledge base in real time based on the target question and the second reply information. When the update method is a batch update method, the first target object can batch enter new question and answer pairs in the target knowledge base. In some embodiments, after the target knowledge base is updated, the offline knowledge base can also be updated synchronously.

[0091] It can be seen that the solution provided in this application achieves the purpose of combining intelligent customer service and manual customer service to answer user questions, thereby achieving the technical effect of improving answering efficiency, and further solving the technical problem of relying on manual customer service to answer user questions in related technologies, resulting in low answering efficiency.

[0092] Example 2

[0093] According to an embodiment of the present invention, an embodiment of a problem handling device is provided, wherein: Fig.11 is a schematic diagram of an optional problem handling device according to an embodiment of the present invention, such as Fig.11 As shown, the device comprises:

[0094] The retrieval module 1101 is used to search the target knowledge base based on the target question to obtain the retrieval result when receiving the target question of the user, wherein the target knowledge base includes a plurality of question-answer pairs, and the question-answer pairs include questions and answer information corresponding to the questions;

[0095] A first determination module 1102 is used for inputting the at least one answer information into the target neural network model when the search result includes at least one answer information, and determining the first answer information for the target question through the target neural network model;

[0096] The second determination module 1103 is used to determine the manual customer service corresponding to the user when the search result does not include the answer information, transfer the user to the manual customer service, and send a first prompt information to the manual customer service, wherein the first prompt information is used to prompt the manual customer service to answer the target question;

[0097] The first sending module 1104 is used to generate second prompt information according to the target work order when a target work order triggered by manual customer service is detected, and send the second prompt information to the first target object, wherein the target work order includes a target question, and the second prompt information is used to prompt the first target object to answer the target question.

[0098] It should be noted that the above-mentioned retrieval module 1101, the first determination module 1102, the second determination module 1103 and the first sending module 1104 correspond to steps S101 to S104 in the above-mentioned embodiment, and the examples and application scenarios implemented by the four modules and the corresponding steps are the same, but are not limited to the contents disclosed in the above-mentioned embodiment 1.

[0099] Optionally, the first sending module further includes: a determination submodule, used to determine the first target object according to the information of the object to receive the target work order in the target work order; and a sending submodule, used to send the second prompt information to the first target object.

[0100] Optionally, the problem handling device also includes: a second sending module, used to receive the work order processing results fed back by the first target object, and send the work order processing results to manual customer service; an updating module, used to extract second reply information for the target problem from the work order processing results when receiving a request to update the knowledge base sent by the manual customer service, and update the target knowledge base based on the target problem and the second reply information.

[0101] Optionally, the problem handling device also includes: a third sending module, which is used to determine the second target object of the target work order to be transferred from the work order transfer request when receiving a work order transfer request sent by the first target object, generate a third prompt message based on the target work order, and send the third prompt message to the second target object, wherein the third prompt message is used to prompt the second target object to answer the target question; a fourth sending module, which is used to generate a fourth prompt message based on the work order information of the target work order when receiving a work order cancellation request sent by manual customer service, and send the fourth prompt message to the first target object, wherein the fourth prompt message is used to prompt the first target object to cancel the processing of the target work order; a fifth sending module, which is used to determine the time point of sending the second prompt message as the target time point, and calculate the first time length between the current time point and the target time point, and when the first time length meets the preset time length condition, send the second prompt message to the first target object again.

[0102] Optionally, the problem handling device also includes: a sixth sending module, which is used to send first information to the user when receiving target information sent by the user, wherein the target information is information that does not include a question, and the first information is used to prompt the user to select the user's subject type; a seventh sending module, which is used to receive the subject type fed back by the user based on the first information, and determine multiple question types corresponding to the subject type, generate second information based on the multiple question types, and send the second information to the user, wherein the second information is used to prompt the user to select the question type; an eighth sending module, which is used to receive the question type fed back by the user based on the second information, and determine multiple candidate questions corresponding to the question type, generate third information based on the multiple candidate questions, and send the third information to the user, wherein the third information is used to prompt the user to select a candidate question; a third determination module, which is used to receive the question fed back by the user based on the third information, and determine the question as the target question.

[0103] Optionally, the retrieval module further includes: a retrieval submodule, which is used to search in the target knowledge base based on the target question to obtain a retrieval result when the target question does not belong to multiple candidate questions.

[0104] Optionally, the question processing device also includes: a fourth determination module, which is used to determine the preset answer information corresponding to the candidate question that is the same as the target question as the answer information for the target question when the target question belongs to multiple candidate questions.

[0105] Example 3

[0106] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is further provided, in which a computer program is stored, wherein the computer program is configured to execute the above-mentioned problem handling method when running.

[0107] Example 4

[0108] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, wherein: Fig.12 is a schematic diagram of an optional electronic device according to an embodiment of the present invention, such as Fig.12 As shown, the electronic device includes one or more processors; a memory for storing one or more programs, which, when the one or more programs are executed by the one or more processors, enables the one or more processors to run the programs, wherein the programs are configured to execute the above-mentioned problem handling method when running.

[0109] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0110] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0111] In the several embodiments provided in this 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 schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. 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 is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0112] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0113] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, 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 software functional units.

[0114] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0115] The above are only preferred embodiments of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A problem solving method, characterized in that: include: When receiving a target question from a user, searching a target knowledge base based on the target question to obtain a search result, wherein the target knowledge base includes a plurality of question-answer pairs, and the question-answer pairs include questions and answer information corresponding to the questions; In the case where the search result includes at least one answer information, the at least one answer information is input into the target neural network model, and the first answer information for the target question is determined by the target neural network model; If the search result does not include answer information, determine the manual customer service corresponding to the user, transfer the user to the manual customer service, and send a first prompt message to the manual customer service, wherein the first prompt message is used to prompt the manual customer service to answer the target question; When a target work order triggered by the manual customer service is detected, second prompt information is generated according to the target work order, and the second prompt information is sent to the first target object, wherein the target work order includes the target question, and the second prompt information is used to prompt the first target object to answer the target question.

2. The method according to claim 1, characterized in that The target work order also includes information of an object to receive the target work order, wherein sending the second prompt information to the first target object includes: Determine the first target object according to information of the object to receive the target work order in the target work order; The second prompt information is sent to the first target object.

3. The method according to claim 1, characterized in that After sending the second prompt information to the first target object, the method further includes: Receive the work order processing result fed back by the first target object, and send the work order processing result to the manual customer service; When receiving the knowledge base update request sent by the manual customer service, second reply information for the target question is extracted from the work order processing result, and the target knowledge base is updated based on the target question and the second reply information.

4. The method according to claim 1, characterized in that After sending the second prompt information to the first target object, the method further includes: Upon receiving a work order transfer request sent by the first target object, determining a second target object to which the target work order is to be transferred from the work order transfer request, generating third prompt information according to the target work order, and sending the third prompt information to the second target object, wherein the third prompt information is used to prompt the second target object to answer the target question; Upon receiving a work order cancellation request sent by the manual customer service, generating fourth prompt information according to the work order information of the target work order, and sending the fourth prompt information to the first target object, wherein the fourth prompt information is used to prompt the first target object to cancel processing of the target work order; The time point for sending the second prompt information is determined as the target time point, and the first time length between the current time point and the target time point is calculated. When the first time length meets the preset time length condition, the second prompt information is sent to the first target object again.

5. The method according to claim 1, characterized in that Before searching the target knowledge base based on the target question to obtain the search results, the method further includes: In case of receiving the target information sent by the user, sending first information to the user, wherein the target information is information not including questions, and the first information is used to prompt the user to select the subject type of the user; receiving a subject type fed back by the user based on the first information, determining a plurality of question types corresponding to the subject type, generating second information according to the plurality of question types, and sending the second information to the user, wherein the second information is used to prompt the user to select a question type; receiving a question type fed back by the user based on the second information, determining a plurality of candidate questions corresponding to the question type, generating third information according to the plurality of candidate questions, and sending the third information to the user, wherein the third information is used to prompt the user to select a candidate question; A question fed back by the user based on the third information is received, and the question is determined as the target question.

6. The method according to claim 5, characterized in that Based on the target question, a search is performed in the target knowledge base to obtain search results, including: In the case that the target question does not belong to the multiple candidate questions, a search is performed in the target knowledge base based on the target question to obtain the search result.

7. The method according to claim 5, characterized in that After determining the problem as the target problem, the method further includes: In the case that the target question belongs to the multiple candidate questions, the preset answer information corresponding to the candidate question that is the same as the target question is determined as the answer information for the target question.

8. A problem handling device, characterized in that: include: A retrieval module is used to, upon receiving a target question from a user, search a target knowledge base based on the target question to obtain a retrieval result, wherein the target knowledge base includes a plurality of question-answer pairs, each of which includes a question and answer information corresponding to the question; A first determination module is used for inputting the at least one answer information into a target neural network model when the search result includes at least one answer information, and determining the first answer information for the target question through the target neural network model; A second determination module is used to determine the manual customer service corresponding to the user when the search result does not include the answer information, transfer the user to the manual customer service, and send a first prompt information to the manual customer service, wherein the first prompt information is used to prompt the manual customer service to answer the target question; The first sending module is used to generate second prompt information according to the target work order when a target work order triggered by the manual customer service is detected, and send the second prompt information to the first target object, wherein the target work order includes the target question, and the second prompt information is used to prompt the first target object to answer the target question.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program is configured to execute the problem handling method described in any one of claims 1 to 7 when running.

10. An electronic device, characterized in that: The electronic device includes one or more processors; A memory for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to run the programs, wherein the programs are configured to execute the problem handling method described in any one of claims 1 to 7 when run.

Images