Intelligent operation and maintenance platform construction method and system based on artificial intelligence

Abstract

The invention provides an intelligent operation and maintenance platform construction method and system based on artificial intelligence, and the method comprises the following steps: analyzing the intention of a user, and obtaining a learning sample and a problem sample; constructing a large language model; and constructing a knowledge base. According to the technical scheme, an intelligent operation and maintenance service platform is constructed, and a positive effect is played in the aspect of improving efficiency; consultation items of daily operation and maintenance personnel are directly solved through the automatic question and answer system, so that the response speed is increased, and the work pressure of an operation and maintenance team is effectively relieved.

Description

Background Art

[0002] The Information Technology Center operates and maintains various business management systems for business back-office users, and undertakes various online business functions such as underwriting, preservation, correction, payment, and claims settlement. During the production and operation process, application system users may encounter operational questions, unclear processes, business interruptions, data errors, unclear management policies or business rules, etc., and such situations are currently handled by the operation and maintenance personnel of the Information Technology Center through the online operation and maintenance system and the problem ticket system.

[0003] The prior art objectively has the following disadvantages:

[0004] 1) Users change frequently and their levels vary widely, making it difficult to achieve system promotion effects through training.

[0005] 2) The questions are highly homogeneous, and similar categories of questions are asked repeatedly.

[0006] 3) In the current form of operation and maintenance problem handling, users often need to queue up to wait for the problem to be handled, and the density of operation and maintenance work in daily work is often uneven. As a result, when problems break out or business peaks, operation and maintenance personnel need to queue up to solve the problems.

[0007] 4) The operation and maintenance work is repetitive and boring, which affects the retention and work enthusiasm of operation and maintenance service personnel, and affects the growth of operation and maintenance personnel. Summary of the invention

[0008] The present application provides a method and system for building an intelligent operation and maintenance platform based on artificial intelligence, which is used to build an intelligent operation and maintenance service platform. The consulting matters of daily operation and maintenance personnel can be directly solved through an automated question-and-answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team.

[0009] In the first aspect, a method for building an intelligent operation and maintenance platform based on artificial intelligence is provided, comprising the following steps:

[0010] Analyze user intentions to obtain learning samples and question samples;

[0011] Build a large language model;

[0012] Build a knowledge base.

[0013] In the above technical solution, learning samples and question samples are obtained by analyzing user intentions; a large language model is built; a knowledge base is built; and an intelligent operation and maintenance service platform is built, which plays a positive role in improving efficiency. The consulting matters of daily operation and maintenance personnel are directly resolved through an automated question-and-answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team.

[0014] In a specific implementation scheme, the steps of building a large language model include:

[0015] The large language model is trained using the learning samples; the problem samples are labeled using the trained large language model to obtain labeled samples; and the large language model is used to perform intent learning.

[0016] In a specific embodiment, it also includes:

[0017] Build interfaces with business systems.

[0018] In a specific embodiment, it also includes:

[0019] The service capability of the interface is constructed by utilizing the large language model and the knowledge base.

[0020] In a specific embodiment, it also includes:

[0021] Build an authority control system.

[0022] In a specific embodiment, it also includes:

[0023] Report user issues;

[0024] Restore the session process;

[0025] Store and analyze question and answer data.

[0026] In a specific implementation scheme, the steps of building a knowledge base include:

[0027] Import existing Q&A scenarios or documents;

[0028] Create a shared document to save the questions and responses from the daily problem handling system, and enter those that do not exist in the question and answer database in a timely manner;

[0029] Update the Q&A library and document library.

[0030] In a specific implementation plan, the steps of building a permission control system specifically include:

[0031] Set query permissions;

[0032] Set up temporary authorization;

[0033] Verify the queried data.

[0034] In a specific implementation scheme, the steps of analyzing the user intention and obtaining the learning samples and question samples specifically include:

[0035] Sort out user questions and collect the intent of the questions;

[0036] Classify user questions and obtain learning samples and question samples.

[0037] Secondly, an intelligent operation and maintenance platform construction system based on artificial intelligence is provided, including:

[0038] The interface module is used to build an interface with the business system and communicate with the business system;

[0039] The large model module is used to analyze user intent, obtain learning samples and question samples, and build a large language model;

[0040] Knowledge base module, used to build a knowledge base;

[0041] A service capability module, used to construct the service capability of the interface by using the large language model and the knowledge base;

[0042] Control system module, used to build the authority control system;

[0043] Problem reporting module, used to report user problems;

[0044] The session restoration module is used to restore the session process;

[0045] The data storage module is used to store and analyze question and answer data.

[0046] In the above technical solution, an interface module is used to build an interface with the business system and communicate with the business system; a large model module is used to analyze user intentions, obtain learning samples and question samples, and build a large language model; a knowledge base module is used to build a knowledge base; a service capability module is used to use the large language model and the knowledge base to build the service capability of the interface; a control system module is used to build an authority control system; a problem reporting module is used to report user problems; a session restoration module is used to restore the session process; a data storage module is used to store and analyze question and answer data; an intelligent operation and maintenance service platform is built, which plays a positive role in improving efficiency; daily operation and maintenance personnel's consulting matters are directly resolved through an automated question and answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flowchart of a method for building an intelligent operation and maintenance platform based on artificial intelligence provided in an embodiment of the present application;

[0048] Figure 2A structural block diagram of an intelligent operation and maintenance platform construction system based on artificial intelligence provided in an embodiment of the present application;

[0049] Figure 3 A task-based session including permission verification flow chart provided in an embodiment of the present application;

[0050] Figure 4 A schematic diagram of a knowledge base provided for an embodiment of the present application;

[0051] Figure 5 A flowchart corresponding to a process session drawn after sorting out the interface problems provided in the embodiment of the present application;

[0052] Figure 6 A schematic diagram of a customer service robot chat provided in an embodiment of the present application;

[0053] Figure 7 Schematic diagram of intelligent capture of chat records by a customer service robot provided in an embodiment of the present application. DETAILED DESCRIPTION

[0054] The present application is further described in detail below through the accompanying drawings and embodiments. Through these descriptions, the characteristics and advantages of the present application will become clearer and more specific.

[0055] The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration." Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. Although various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless otherwise noted.

[0056] In addition, the technical features involved in different embodiments of the present application described below can be combined with each other as long as they do not conflict with each other.

[0057] To facilitate understanding of the method and system for building an intelligent operation and maintenance platform based on artificial intelligence provided by the embodiment of the present application, its application scenario is first explained. The method and system for building an intelligent operation and maintenance platform based on artificial intelligence provided by the embodiment of the present application are used to build an intelligent operation and maintenance service platform. The consultation matters of daily operation and maintenance personnel can be directly solved through an automated question-and-answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team. The information technology center operates and maintains various business management systems for business back-office users, and undertakes various business online functions such as underwriting, preservation, correction, payment, and claims settlement. During the production and operation process, application system users will encounter operational questions, unclear processes, business interruptions, data errors, unclear management policies or business rules, etc. during the use of the system. Such situations are currently handled by the operation and maintenance personnel of the information technology center through the online operation and maintenance system and the problem work order system. The existing technology objectively has the following disadvantages: 1) Users change frequently, and the user levels are uneven, so it is difficult to achieve the promotion effect of system use through training. 2) The problems are highly homogeneous, and similar types of problems are repeatedly raised. 3) In the current form of operation and maintenance problem handling, users often need to queue up and wait for problems to be handled, while the density of operation and maintenance work in daily work is often uneven. As a result, when problems break out or business peaks, operation and maintenance personnel need to arrange queues to solve the problems. 4) Operation and maintenance work is repetitive and boring, which affects the retention and work enthusiasm of operation and maintenance service personnel, and affects the growth of operation and maintenance personnel. For this reason, the embodiment of the present application provides a method and system for building an intelligent operation and maintenance platform based on artificial intelligence to build an intelligent operation and maintenance service platform. The consultation matters of daily operation and maintenance personnel can be directly solved through an automated question-and-answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team. It is described in detail below with specific drawings and embodiments.

[0058] refer to Figures 1 to 7 , Figure 1 A flowchart of a method for building an intelligent operation and maintenance platform based on artificial intelligence provided in an embodiment of the present application; Figure 2 A structural block diagram of an intelligent operation and maintenance platform construction system based on artificial intelligence provided in an embodiment of the present application; Figure 3 A task-based session including permission verification flow chart provided in an embodiment of the present application; Figure 4 A schematic diagram of a knowledge base provided for an embodiment of the present application; Figure 5 A flowchart corresponding to a process session drawn after sorting out the interface problems provided in the embodiment of the present application; Figure 6 A schematic diagram of a customer service robot chat provided in an embodiment of the present application; Figure 7 Schematic diagram of intelligent capture of chat records by a customer service robot provided in an embodiment of the present application.

[0059] exist Figure 1In the embodiment of the present application, a method for building an intelligent operation and maintenance platform based on artificial intelligence is provided, comprising the following steps:

[0060] Analyze user intentions to obtain learning samples and question samples;

[0061] Build a large language model;

[0062] Build a knowledge base.

[0063] In the above technical solution, learning samples and question samples are obtained by analyzing user intentions; a large language model is built; a knowledge base is built; and an intelligent operation and maintenance service platform is built, which plays a positive role in improving efficiency. The consulting matters of daily operation and maintenance personnel are directly resolved through an automated question-and-answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team.

[0064] In a specific implementation scheme, the steps of building a large language model include:

[0065] The large language model is trained using the learning samples; the problem samples are labeled using the trained large language model to obtain labeled samples; and the large language model is used to perform intent learning.

[0066] In a specific embodiment, it also includes:

[0067] Build interfaces with business systems.

[0068] In a specific embodiment, it also includes:

[0069] The service capability of the interface is constructed by utilizing the large language model and the knowledge base.

[0070] In a specific embodiment, it also includes:

[0071] Build an authority control system.

[0072] In a specific embodiment, it also includes:

[0073] Report user issues;

[0074] Restore the session process;

[0075] Store and analyze question and answer data.

[0076] In a specific implementation scheme, the steps of building a knowledge base include:

[0077] Import existing Q&A scenarios or documents;

[0078] Create a shared document to save the questions and responses from the daily problem handling system, and enter those that do not exist in the question and answer database in a timely manner;

[0079] Update the Q&A library and document library.

[0080] In a specific implementation plan, the steps of building a permission control system specifically include:

[0081] Set query permissions;

[0082] Set up temporary authorization;

[0083] Verify the queried data.

[0084] In a specific implementation scheme, the steps of analyzing the user intention and obtaining the learning samples and question samples specifically include:

[0085] Sort out user questions and collect the intent of the questions;

[0086] Classify user questions and obtain learning samples and question samples.

[0087] refer to Figures 3 to 7 Specifically, the specific steps of the method for building an intelligent operation and maintenance platform based on artificial intelligence include:

[0088] (I) Session entry (i.e. building an interface with the business system)

[0089] Each business system integrates the intelligent operation and maintenance service platform, which requires the business system to integrate the relevant components of intelligent operation and maintenance and open the platform conversation interface for dialogue.

[0090] 2. User Intention Analysis and User Question and Answer Retrieval

[0091] Problem sorting: Extract the system's operation and maintenance problems for a month, classify and label the problems. Considering the complexity of the problem, first limit the scope to problems that can be solved within 1 hour. Such problems currently account for about 50%-70% of all lines. During the sorting process, collect the intention of the problem, the information that needs to be provided by the user, similar questions, frequency of occurrence, etc.

[0092] Problem classification: one is interface type, the other is question and answer type; interface type selection reference is the problem with high frequency and clear operation and maintenance processing process. For example: the insurance application has been paid, but the insurance policy has not been generated. The question and answer type will reply to the user with some common problem handling methods, such as the transfer modification real-name authentication cannot pass, the expired agency fee cannot be processed, and the non-auto insurance business needs to backdate the bill. This type of problem depends on the nature of the problem. There is no need for operation and maintenance to intervene to help the user handle it, but the reason needs to be explained to the user.

[0093] Sample labeling: Accurately classify the 200 questions, label their intent, ensure their correctness, and use them as samples for machine learning to train the model; extract enough question samples, use the trained model to label the samples, and the labeled samples can be provided to the model for intent learning.

[0094] 3. Knowledge base construction

[0095] refer to Figure 4 , with the help of the large model RAG capability (Retrieval-Augmented Generation, i.e. retrieval enhanced generation) to achieve document retrieval, slicing and other related functions; on the one hand, give priority to entering previously existing question and answer scenarios or documents; on the other hand, establish a shared document to retain the questions and replies of the daily question handling system, and enter them in time when they do not exist in the question and answer library. At the same time, if there are any process changes, the question and answer library and document library will be updated synchronously after the upgrade is launched to ensure the comprehensiveness and timeliness of the knowledge base content.

[0096] (IV) Interface service capabilities

[0097] refer to Figures 5 to 7 After sorting out the interface-related issues, draw a flowchart corresponding to the process conversation and the corresponding reply script. This process covers all scenarios of the entire intention, that is, after calling the interface, all corresponding related states can be viewed.

[0098] (V) Construction of the authority system

[0099] Data security is of utmost importance in the insurance industry. To prevent data leakage risks, a permission control system has been established. First, query the query permission according to the current session + interface identifier + personnel + business system code + other identifiers. When connecting the interface, check whether there is permission first; second, because the operation and maintenance issues of each business system are strongly related, temporary authorization needs to be set when querying the permissions of two systems in one process; third, the queried data needs to be verified whether it needs to be desensitized, etc., and it can only be displayed after all are passed.

[0100] (VI) One-click reporting of issues

[0101] If the conversation process of the intelligent operation and maintenance platform does not solve the user's related problems, you can initiate a "one-click report" in the conversation to quickly report the user's problem to the problem service desk managed by the operation and maintenance personnel.

[0102] (VII) Session Process Restoration

[0103] When reporting a problem to a manually handled service desk, the platform restores the previous conversation with the user and adds an attachment to the reported problem, allowing the processing staff to quickly understand what happened without the user having to describe the problem a second time.

[0104] (VIII) Question and answer data storage and analysis

[0105] Currently, the conversation content platform with users has realized the storage function, analyzes user query and interaction data, and provides management with insights into system performance, user needs and potential problems, thereby supporting more accurate business decisions.

[0106] In the above technical solutions, the application of the constructed intelligent operation and maintenance platform has significant effects in the following aspects:

[0107] 1. Liberate manpower

[0108] The application of the intelligent operation and maintenance service platform has achieved a 20% decrease in the number of problem tickets, which shows that the platform has played a positive role in improving efficiency. Daily operation and maintenance personnel's consultation matters can be directly resolved through the automated question-and-answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team, allowing human resources to focus on more complex and higher-value work.

[0109] 2. Assist business development

[0110] By storing and analyzing user conversations, we can gain deep insights into application system performance, user needs, and potential problems, which is important for management to make accurate business decisions. At the same time, data analysis results can guide us to optimize product functions, improve user experience, and prevent possible problems in advance, which also reflects our scientific management method of using data-driven decision-making in the process of digital transformation.

[0111] 3. Enhance user experience

[0112] Providing instant feedback and solutions increases user satisfaction and dependence on the system and improves the overall user experience.

[0113] 4. Reduce adverse effects and improve customer retention

[0114] Provide quick solutions to the business impacts of the company's customers caused by insufficient skills of back-office staff or system interruptions, reduce the adverse impact on customer business handling, reduce or avoid the loss of potential customers, reduce user complaints, and reduce the negative impact on the company's image

[0115] 5. Continuous knowledge updating and optimization

[0116] Through the continuous use and feedback of intelligent operation and maintenance services, the knowledge base can be continuously optimized and updated, ensuring the synchronous update of solutions and business processes, and improving the accuracy and adaptability of services.

[0117] Furthermore, the intelligent operation and maintenance platform we built uses advanced natural language understanding algorithms to more accurately capture and parse user descriptions in different contexts, even if these descriptions deviate from standard terms. In addition, the artificial intelligence model integrates reasoning capabilities to enhance the response effect to complex queries in professional fields. Specific operational problems that arise on the service platform can be efficiently and accurately supported by our model, injecting new impetus into the solution. Figure 3 , including:

[0118] 1. User Intent Identification

[0119] It is very important to classify and label the large amount of user question data, operation and maintenance session records, and work order reports collected, which provides valuable training materials for subsequent machine learning. Based on these historical data samples, combined with advanced machine learning algorithms and natural language processing technology, we train more accurate prediction models to more effectively understand the user's true intentions, thereby improving the accuracy and efficiency of our questions to users. Using this data-driven methodology, the intelligent operation and maintenance platform can quickly respond to user needs and solve problems immediately.

[0120] (II) Three-layer layout implementation

[0121] Question-and-answer dialogue implementation: The conversational robot is equipped with powerful process processing capabilities, and it can perform a series of operations based on the user's specific intentions. During the interaction process, the robot can not only initiate questions, but also identify and extract user information by calling the language model API, which helps us understand the problems faced by users more accurately.

[0122] Knowledge base construction: Build a knowledge system with rich content, reasonable structure and easy retrieval, which can store and organize a large amount of business knowledge information, including FAQs, business process guidance, policy regulations, etc., to support intelligent assistants and service personnel to provide professional customer service efficiently. Through artificial intelligence technology and data analysis, we can also learn and optimize customer interactions to continuously improve the quality of the knowledge base and service performance.

[0123] Problem Solving: Through the intelligent operation and maintenance platform, we can get close to front-line business personnel and get in touch with daily business problems and various operational problems. After the problem is accurately located, the process session can call a specific business interface to handle the operation for the problem. Whether the user's needs are to query information, restore business, or connect to the subsequent business, the platform is already capable of providing corresponding services.

[0124] exist Figure 2 In the embodiment of the present application, a system for building an intelligent operation and maintenance platform based on artificial intelligence is provided, including:

[0125] The interface module is used to build an interface with the business system and communicate with the business system;

[0126] The large model module is used to analyze user intent, obtain learning samples and question samples, and build a large language model;

[0127] Knowledge base module, used to build a knowledge base;

[0128] A service capability module, used to construct the service capability of the interface by using the large language model and the knowledge base;

[0129] Control system module, used to build the authority control system;

[0130] Problem reporting module, used to report user problems;

[0131] The session restoration module is used to restore the session process;

[0132] The data storage module is used to store and analyze question and answer data.

[0133] In the above technical solution, an interface module is used to build an interface with the business system and communicate with the business system; a large model module is used to analyze user intentions, obtain learning samples and question samples, and build a large language model; a knowledge base module is used to build a knowledge base; a service capability module is used to use the large language model and the knowledge base to build the service capability of the interface; a control system module is used to build an authority control system; a problem reporting module is used to report user problems; a session restoration module is used to restore the session process; a data storage module is used to store and analyze question and answer data; an intelligent operation and maintenance service platform is built, which plays a positive role in improving efficiency; daily operation and maintenance personnel's consulting matters are directly resolved through an automated question and answer system, which not only improves the response speed, but also effectively reduces the work pressure of the operation and maintenance team.

[0134] Those skilled in the art will appreciate that the present application may be implemented as a system, method or computer program product.

[0135] Therefore, the present disclosure may be specifically implemented in the following forms, namely: it may be completely hardware, it may be completely software (including firmware, resident software, microcode, etc.), or it may be a combination of hardware and software, generally referred to herein as a "circuit", "module" or "system". In addition, in some embodiments, the present application may also be implemented in the form of a computer program product in one or more computer-readable media, and the computer-readable medium may contain computer-readable program code.

[0136] Any combination of one or more computer-readable media can be used. Computer-readable media can be computer-readable signal media or computer-readable storage media. Computer-readable storage media can be, for example, but not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or devices, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: electrical connections with one or more wires, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this document, computer-readable storage media can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0137] Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and cannot be understood as limiting the present application. A person of ordinary skill in the art can change, modify, replace and modify the above embodiments within the scope of the present application. On this basis, a variety of replacements and improvements can be made to the present application, all of which fall within the scope of protection of the present application.

Claims

1. A method for constructing an intelligent operation and maintenance platform based on artificial intelligence, characterized in that: The following steps are involved: Analyze user intentions to obtain learning samples and question samples; Build large language models; Build a knowledge base.

2. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 1, characterized in that: The steps to build a large language model include: The large language model is trained using the learning samples; the problem samples are labeled using the trained large language model to obtain labeled samples; and the large language model is used to perform intent learning.

3. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 2, characterized in that: Also includes: Build interfaces with business systems.

4. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 3, characterized in that: Also includes: The service capability of the interface is constructed by utilizing the large language model and the knowledge base.

5. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 4, characterized in that: Also includes: Build an authority control system.

6. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 5, characterized in that: Also includes: Report user issues; Restore the session process; Store and analyze question and answer data.

7. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 6, characterized in that: The steps to build a knowledge base include: Import existing Q&A scenarios or documents; Create a shared document to save the questions and responses from the daily problem handling system, and enter those that do not exist in the question and answer database in a timely manner; Update the Q&A library and document library.

8. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 7, characterized in that: The steps to build a permission control system include: Set query permissions; Set up temporary authorization; Verify the queried data.

9. The method for constructing an intelligent operation and maintenance platform based on artificial intelligence according to claim 8, characterized in that: The steps of analyzing user intentions and obtaining learning samples and question samples include: Sort out user questions and collect the intent of the questions; Classify user questions and obtain learning samples and question samples.

10. An intelligent operation and maintenance platform construction system based on artificial intelligence, characterized in that: include: The interface module is used to build an interface with the business system and communicate with the business system; The large model module is used to analyze user intent, obtain learning samples and question samples, and build a large language model; Knowledge base module, used to build a knowledge base; A service capability module, used to construct the service capability of the interface by using the large language model and the knowledge base; Control system module, used to build the authority control system; Problem reporting module, used to report user problems; The session restoration module is used to restore the session process; The data storage module is used to store and analyze question and answer data.

Images