Optimization scheme determination methods, devices, storage media and program products

Abstract

This application provides an optimization scheme determination method, apparatus, storage medium, and program product, relating to the field of wireless communication technology. It can improve the detection accuracy of root causes of problems, generate refined network optimization schemes, and improve the efficiency of network optimization work. The method includes: acquiring a complaint ticket sent by a terminal, determining, based on the complaint ticket, the target root cause causing the terminal's performance to be in a state requiring optimization, and then generating a target optimization scheme based on the target root cause. The complaint ticket is used to identify that the terminal's performance is in a state requiring optimization, and the target optimization scheme is used to improve the terminal's performance.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a method, apparatus, storage medium, and program product for determining an optimization scheme. Background Technology

[0002] With the rapid development of mobile network technology, the number of devices accessing mobile networks continues to grow, and users' demand for data transmission is also experiencing explosive growth, which places higher demands on the stability of mobile networks. Against this backdrop, efficient mobile network optimization and operation has become a critical issue that operators urgently need to address.

[0003] However, existing network optimization and operation systems have certain limitations. Specifically, while current network optimization technologies can automatically detect problems in the network, they lack further analysis and decision-making capabilities and cannot automatically provide specific optimization suggestions. This forces operators to delegate work orders layer by layer to frontline personnel after identifying problems. However, frontline personnel often have weaker skill levels and lack user perception analysis and root cause identification tools. This makes it difficult to perform refined operations when handling network optimization tasks, often resulting in only some relatively crude optimization measures. Consequently, network optimization work is inefficient and fails to meet users' demands for high-quality mobile network services. Summary of the Invention

[0004] This application provides a method, apparatus, storage medium, and program product for determining optimization schemes, which can improve the detection accuracy of root causes of problems, generate refined and targeted network optimization schemes, and thus improve the efficiency of network optimization work.

[0005] To achieve the above objectives, this application adopts the following technical solution:

[0006] Firstly, this application provides a method for determining an optimization scheme, applied to a server, the method comprising:

[0007] Obtain the complaint tickets sent by the terminal, and based on the complaint tickets, determine the root cause that causes the terminal's performance to be in a state of unoptimization, and then generate a target optimization plan based on the target root cause.

[0008] Among them, the complaint ticket is used to indicate that the performance of the terminal is in a state of needing optimization, and the target optimization plan is used to improve the performance of the terminal.

[0009] Based on the above solution, the server can accurately pinpoint the root cause of the terminal's unoptimized state using information from the complaint ticket, and generate an optimization plan to improve terminal performance based on this root cause. This provides a basis for subsequent optimization work, improves network optimization efficiency, and ultimately meets users' demands for high-quality mobile network services.

[0010] Optionally, the above-mentioned determination of the root cause of the terminal's performance being in a state of unoptimized performance based on the complaint ticket may specifically include: identifying the network device associated with the terminal's performance, then performing fault detection on the network device, obtaining the detection results, and, if the detection results indicate that the network device has failed, determining the root cause of the network device failure, and then using the root cause as the target root cause of the terminal's performance being in a state of unoptimized performance.

[0011] Optionally, the root cause of the fault can be characterized by a failure in the first component of the network device, and the target optimization solution is to repair the first component.

[0012] Optionally, the server may store optimization solutions corresponding to multiple root causes, including a target root cause. Based on this, the above-mentioned generation of a target optimization solution based on the target root cause may specifically include: determining at least one optimization solution corresponding to the target root cause, and then determining the target optimization solution based on the at least one optimization solution corresponding to the target root cause.

[0013] Optionally, the server can be deployed with an information extraction model. Based on this model, optimization solutions corresponding to multiple root causes can be obtained in the following way: obtain textual information, and then extract the optimization solution corresponding to each root cause from the textual information using the information extraction model.

[0014] The textual information contains multiple root causes and the optimization scheme corresponding to each root cause.

[0015] Optionally, the information extraction model can be trained in the following way: obtain training data that identifies optimization schemes corresponding to different root causes, and use the training data to train a large language model to obtain the information extraction model.

[0016] Optionally, the server can be deployed with a root cause analysis model and a solution generation model. The root cause analysis model is used to identify the target root cause causing the terminal's performance to be in a state of unoptimized performance, based on complaint tickets. The solution generation model is used to generate target optimization solutions based on the target root cause.

[0017] Secondly, this application provides an optimization scheme determination apparatus, comprising:

[0018] The transceiver unit is used to obtain complaint work orders sent by the terminal.

[0019] The determination unit is used to identify the root causes that cause the terminal's performance to be in a state of unoptimization, based on the complaint work order.

[0020] The processing unit is used to generate an optimization scheme for the target based on the root cause of the target.

[0021] Optionally, the determining unit is also used to determine the network device associated with the performance of the terminal.

[0022] Optionally, the processing unit is also used to perform fault detection on network devices and obtain detection results.

[0023] Optionally, the determining unit is also configured to, when the detection results indicate that the network device has failed, determine the root cause of the failure and use the root cause as the target root cause that leads to the terminal's performance being in a state of unoptimization.

[0024] Optionally, the determining unit is also used to determine at least one optimization scheme corresponding to the target root cause.

[0025] Optionally, the determining unit is further configured to determine a target optimization scheme based on at least one optimization scheme corresponding to the target root cause.

[0026] Optionally, the processing unit is also used to extract the optimization scheme corresponding to each root cause from the textual information through an information extraction model.

[0027] Optionally, the processing unit is also used to train the large language model using training data to obtain an information extraction model.

[0028] Thirdly, this application provides an optimization scheme determination apparatus, which includes a processor and a communication interface. The communication interface is coupled to the processor, and the processor is used to run computer programs or instructions to implement the optimization scheme determination method as described in any one of the first aspects and any possible implementations of the first aspect.

[0029] Fourthly, this application provides a computer-readable storage medium storing instructions that, when executed on a terminal, cause the terminal to perform the optimization scheme determination method as described in any one of the first aspects and any possible implementations of the first aspect.

[0030] Fifthly, this application provides a computer program product comprising computer instructions that, when executed on a computer, cause the computer to perform the optimization scheme determination method as described in any one of the first aspects and any possible implementations of the first aspect.

[0031] It is understood that the beneficial effects that can be achieved by the second to fifth aspects provided above can be referred to the beneficial effects in any possible design of the optimization scheme determination method as described in any one of the first aspects and any possible implementation of the first aspect, which will not be repeated here. Attached Figure Description

[0032] Figure 1An architecture diagram of an optimization scheme for determining a system is provided in an embodiment of this application;

[0033] Figure 2 A flowchart illustrating an optimization scheme determination method provided in an embodiment of this application;

[0034] Figure 3 A diagram illustrating a knowledge graph construction method provided in an embodiment of this application;

[0035] Figure 4 The knowledge graph provided in this application includes entities and the relationship graph between entities;

[0036] Figure 5 A schematic diagram illustrating the types of large language models provided in embodiments of this application;

[0037] Figure 6 A schematic diagram illustrating the training process of a large language model, provided for an embodiment of this application;

[0038] Figure 7 A schematic diagram illustrating the types of root cause localization models provided in the embodiments of this application;

[0039] Figure 8 An interactive flowchart illustrating an optimization scheme determination method provided in this application embodiment;

[0040] Figure 9 This is a schematic diagram of an optimization scheme determination device provided in an embodiment of this application;

[0041] Figure 10 This is a schematic diagram of the structure of an optimization scheme determination device provided in an embodiment of this application. Detailed Implementation

[0042] The following description, in conjunction with the accompanying drawings, details an optimization scheme determination method and apparatus provided in the embodiments of this application.

[0043] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0044] The terms "first" and "second," etc., used in the specification and drawings of this application are used to distinguish different objects or to distinguish different treatments of the same object, rather than to describe a specific order of objects.

[0045] Furthermore, the terms "comprising" and "having," and any variations thereof, used in the description of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the steps or units listed, but may optionally include other steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus.

[0046] It should be noted that in the embodiments of this application, the words "exemplary" or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design scheme described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design schemes. Specifically, the use of the words "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0047] In recent years, with the rapid development and widespread adoption of 5G technology, mobile networks have become an indispensable infrastructure in modern society, profoundly impacting people's daily lives. Users can instantly access social media information, conduct video communications, browse global news, learn new knowledge, and enjoy convenient services such as mobile payments and online shopping through mobile networks.

[0048] However, with the continuous increase in the number of devices accessing mobile networks, users' demand for data transmission is also showing an explosive growth trend, which places higher demands on the stability of mobile networks and user experience. In order to ensure the service quality, stability, and efficiency of mobile networks, efficient mobile network optimization and operation has become an urgent problem for operators to solve. Effective network optimization can not only improve network performance, but also significantly enhance user satisfaction.

[0049] However, existing network optimization and operation systems have certain limitations when facing these diverse challenges. Specifically, while current network optimization technologies can automatically detect problems in the network, they lack further analysis and decision-making capabilities and cannot automatically provide specific optimization suggestions. This forces operators to delegate work orders layer by layer to frontline personnel after identifying problems. However, frontline personnel often have weaker skill levels and lack user perception analysis and root cause identification tools. This makes it difficult to perform refined operations when handling network optimization tasks, often resulting in only some relatively crude optimization measures. Consequently, network optimization work is inefficient and fails to meet users' demands for high-quality mobile network services.

[0050] To address the aforementioned technical problems, this application provides an optimization scheme determination method. After receiving a complaint ticket from a terminal, the server can accurately pinpoint the root cause causing the terminal to be in an unoptimized state based on the information in the complaint ticket. Then, based on the root cause, it can generate an optimization scheme that can improve terminal performance. This provides a basis for subsequent optimization work, improves network optimization efficiency, and ultimately meets users' needs for high-quality mobile network services.

[0051] Figure 1 An optimization scheme for determining the system architecture diagram provided in this application embodiment is shown below. Figure 1 As shown, the system framework includes: terminal 101 and server 102.

[0052] The terminal 101 may be a device that provides voice and / or data connectivity to a user, a device with wireless connectivity, or other devices connected to a wireless modem. The terminal device may be at least one of a desktop computer, laptop, wireless terminal, or laptop computer. In one embodiment, the electronic device has communication capabilities and can access a wired or wireless network.

[0053] This application embodiment does not limit the number of terminals 101 in the system for determining the optimization scheme, and may include a ratio of... Figure 1 More or fewer terminals 101.

[0054] Server 102 can be a high-performance server providing various services on the internet. It can be a standalone physical server, a server cluster consisting of multiple physical servers, or at least one of the following cloud servers providing basic cloud computing services: cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks, and big data or artificial intelligence platforms. This application embodiment does not limit this. Of course, the server can also include other functions to provide more comprehensive and diversified services.

[0055] In some embodiments, a user holding terminal 101 can submit a complaint ticket to the telecommunications operator through terminal 101 if they intend to improve the performance of terminal 101 (e.g., improve the network connection stability of terminal 101). Accordingly, the user can fill out a complaint ticket in the official application (APP) of the telecommunications operator installed on terminal 101 and trigger a submission operation. In response to the user's submission operation, terminal 101 sends the complaint ticket to server 102 (i.e., the telecommunications operator's server). After receiving the complaint ticket, server 102 can determine the root cause causing the performance of terminal 101 to be in a state of unoptimized performance based on the complaint ticket, and then generate a target optimization plan based on the root cause.

[0056] Figure 2 A flowchart illustrating an optimization scheme determination method provided in this application embodiment is shown below. Figure 2 As shown, the method is composed of Figure 1 The server shown executes a method that includes:

[0057] S201, retrieve the complaint work order sent by the terminal.

[0058] Among them, the complaint ticket is used to indicate that the performance of the terminal is in a state of needing optimization.

[0059] This application does not limit the content included in the complaint ticket. For example, the complaint ticket may include basic terminal information, such as terminal identification (ID) and terminal model. Another example is that the complaint ticket may include complaint content, which includes information related to terminal performance issues, such as unstable network connection, application lag, or prolonged unresponsiveness. Yet another example is that the complaint ticket may include the terminal's location information.

[0060] S202, based on the complaint ticket, identify the root cause that is causing the terminal's performance to be in a state of unoptimization.

[0061] Specifically, the server stores the root causes of multiple terminal performance issues. Accordingly, upon receiving a support ticket, the server can extract information related to terminal performance (such as network connection interruptions) from the support ticket. Then, based on the extracted information, the server can determine the root cause corresponding to the terminal performance-related information from the stored root causes of the multiple terminal performance issues, and use this root cause as the target root cause causing the terminal's performance to be in a state of unoptimized performance.

[0062] In some embodiments, the above-mentioned determination of the target root cause that causes the terminal's performance to be in an unoptimized state based on the complaint work order may specifically include: identifying the network device associated with the terminal's performance, then performing fault detection on the network device, obtaining the detection result, and, if the detection result indicates that the network device has failed, determining the root cause of the network device failure, and using the root cause as the target root cause that causes the terminal's performance to be in an unoptimized state.

[0063] This application does not limit the type of root cause of the fault. For example, the root cause of the fault may be a hardware failure of the network device, which may include power failure, hard drive failure, antenna failure, or damage to physical connections (such as network cables, optical fibers, etc.). As another example, the root cause of the fault may be a software failure of the network device, which may include system conflicts, compatibility issues, data corruption, or external interference such as network attacks or virus infections.

[0064] Specifically, the server can extract the terminal's location information and performance-related information from the complaint ticket. Then, based on the terminal's location information, the server can query the network devices connected to the terminal (such as base stations and switches) within the network topology. Next, based on the terminal's performance-related information, the server can identify the target network device associated with the terminal's performance among the connected network devices and determine a detection plan for the target network device (e.g., detecting the stability of the base station's signal transmission power). Finally, the server can perform detection on the target network device and obtain the detection results.

[0065] If the detection results indicate that the target network device has malfunctioned, the server can determine the root cause of the malfunction based on the detection results and use this root cause as the target root cause leading to the terminal's performance being in a state of unoptimized performance. If the detection results indicate that the target network device has not malfunctioned, the server can also monitor and analyze network traffic using network traffic monitoring tools to determine if network congestion exists. If congestion exists, the server can determine the cause of the network congestion based on the monitoring and analysis results and use this cause as the target root cause leading to the terminal's performance being in a state of unoptimized performance.

[0066] Optionally, after determining the detection plan for the target network device, the server can also send the target network device information and the detection plan to an operating terminal held by a professional. The professional can obtain the target network device information and the detection plan through the operating terminal, and then use professional testing tools to perform fault detection on the target network device according to the detection plan, obtain the detection results, and then send the detection results to the server through the operating terminal.

[0067] S203, based on the root cause of the target, generate an optimization scheme for the target.

[0068] Among them, the target optimization scheme is used to improve the performance of the terminal.

[0069] Specifically, the server stores optimization solutions corresponding to multiple root causes, including a target root cause. Accordingly, the server can determine at least one optimization solution corresponding to the target root cause from the stored optimization solutions for the multiple root causes. Then, the server can select one optimization solution from the at least one optimization solution corresponding to the target root cause as the target optimization solution.

[0070] This application does not limit the method by which the server stores optimization solutions corresponding to multiple root causes. For example, the server can store optimization solutions corresponding to multiple root causes in a database, which may include a graph database, a relational database, and a non-relational database. Another example is that the server can store optimization solutions corresponding to multiple root causes in a file system.

[0071] For example, consider storing optimization solutions corresponding to multiple root causes in a graph database. Accordingly, the server can search the graph database based on the target root cause to obtain at least one optimization solution corresponding to the target root cause. Then, the server can select one optimization solution from the at least one optimization solution corresponding to the target root cause as the target optimization solution.

[0072] In some embodiments, the root cause of the fault can be characterized by a failure in a first component of the network device, and the target optimization scheme can be to repair the first component.

[0073] For example, taking the corrosion of the base station antenna surface as a root cause of the fault, the target optimization scheme generated by the server can clean the corroded part of the antenna and apply anti-corrosion coating to the antenna surface.

[0074] Through the above technical solution, the server can locate network devices associated with the terminal's performance based on information from complaint tickets, perform fault detection on these devices, and then determine the root cause of the terminal's performance being in a state requiring optimization based on the detection results. This improves the accuracy of fault detection and provides a basis for generating subsequent optimization solutions. Subsequently, the server can generate optimization solutions that improve terminal performance based on the root cause of the terminal's unoptimized state. In this way, refined and targeted optimization solutions can be generated, providing a basis for subsequent optimization work, thereby improving the efficiency of network optimization and meeting users' needs for high-quality mobile network services.

[0075] In some embodiments, an information extraction model is deployed in the server. Based on this, in S203 above, the optimization schemes corresponding to multiple root causes stored in the server can be obtained in the following way: the server can obtain textual information, and then extract the optimization scheme corresponding to each root cause from the textual information through the information extraction model.

[0076] The textual information contains multiple root causes and the optimization scheme corresponding to each root cause.

[0077] This application does not limit the type of text information. For example, text information can be an optimization guide, a historical optimization case study, or experience information from network optimization experts in handling network problems.

[0078] Specifically, the server stores text-based information. Correspondingly, the server can input this internally stored text-based information into an information extraction model. The information extraction model can then use natural language processing techniques to extract information from the text-based information, ultimately obtaining the optimization solution corresponding to each root cause output by the information extraction model.

[0079] Through the above scheme, the server can extract optimization solutions corresponding to each root cause from various types of text information through an intelligent information extraction process, thereby providing accurate and effective decision support for network optimization work.

[0080] In some embodiments, after extracting the optimization scheme corresponding to each root cause from textual information using the information extraction model, the method by which the server stores the optimization schemes corresponding to multiple root causes in a graph database includes the following steps:

[0081] (1) Construct the first knowledge graph.

[0082] Specifically, such as Figure 3 As shown, the server can use natural language processing techniques to identify entities in the optimization scheme corresponding to each root cause and extract the relationships between entities. Then, the server can use entities as vertices and the relationships between entities as edges to obtain the first knowledge graph.

[0083] In this embodiment, the method for constructing the first knowledge graph can be based on a knowledge graph construction framework, which includes two modules: a visual editing module and a knowledge graph supplementation module based on multi-turn dialogue. The visual editing module can be used to display entities and relationships between entities in the first knowledge graph, and to provide operations such as editing, updating, and deleting. The knowledge graph supplementation module based on multi-turn dialogue can be used to engage in multi-turn dialogues with the user, automatically extract entities and relationships between entities from the user-input text, and store the extracted entities and relationships into the first knowledge graph.

[0084] (2) Store the entities and relationships between entities in the first knowledge graph into the graph database.

[0085] This application does not limit the type of graph database. For example, the graph database can be a Neo4j graph database, a Redis Graph graph database, or a Janus Graph graph database.

[0086] In some embodiments, the server may store information about root cause-related network devices in a graph database, specifically including the following steps:

[0087] (1) Obtain basic information about network devices.

[0088] This application does not limit the basic information of network devices. For example, the basic information of network devices may include the type of network device, the location information of the network device, and key indicators of the network device, such as performance management (PM) indicators and configuration management (CM) indicators.

[0089] Specifically, network devices can send basic information to the server at preset intervals. Correspondingly, the server can receive the basic information sent by the network devices.

[0090] (2) Preprocess the basic information of the network device to obtain the preprocessed basic information.

[0091] This application does not limit the steps for preprocessing the basic information of network devices in its embodiments. For example, the server can perform integrity checks on the basic information of network devices. Another example is that the server can handle outliers in the basic information of network devices. Yet another example is that the server can perform regularization and normalization processing on the basic information of network devices. Still another example is that the server can perform tagging and bucketing processing on the basic information of network devices.

[0092] For example, consider base stations that include macro base stations and micro base stations. The server can tag the base station type, such as using 0 as the tag for macro base stations and 1 as the tag for micro base stations.

[0093] For example, taking traffic passing through base stations within a week as the basic information, the server can statistically analyze the daily traffic passing through base stations within a week, and then perform bucketing and tagging processing on the statistical traffic. For example, traffic data with a daily total traffic in the range of 0-10 gigabytes (GB) is divided into the first bucket and labeled with 1, and traffic data with a daily total traffic in the range of 10-20GB is divided into the second bucket and labeled with 2.

[0094] (3) Construct a second knowledge graph and store the entities and relationships between entities in the second knowledge graph in a graph database.

[0095] Specifically, the server can refer to the method described above for constructing the first knowledge graph to construct the second knowledge graph, and store the entities and relationships between entities in the second knowledge graph in the graph database.

[0096] In some embodiments, after obtaining the first knowledge graph and the second knowledge graph described above, as follows: Figure 3 As shown, the server can use entity alignment technology to fuse the first and second knowledge graphs to obtain a fused optimized knowledge graph.

[0097] Figure 4 The diagram illustrates the entities contained in the optimized knowledge graph and the relationships between them. Entities include root causes, optimization solutions, reasons, key indicators, work orders (i.e., complaint work orders), antennas, cells, low-sensitivity, base stations, and frontline actions.

[0098] Relationships between entities include: Optimization Solution-Resolve-Root Cause, indicating that the optimization solution can resolve the problem caused by the root cause; Work Order-Corresponding Action-First-Line Action, indicating that the first-line action corresponding to the complaint in the work order (e.g., adjusting the base station antenna); Work Order-Corresponding Indicator-Key Indicator, indicating that the key indicator corresponding to the complaint in the work order (e.g., network latency); Work Order-Corresponding Root Cause-Root Cause, indicating that the root cause corresponding to the complaint in the work order; First-Line Action-Resolve-Root Cause, indicating that the first-line action can resolve the problem caused by the root cause; Antenna-Service-Cell, indicating that the antenna can provide wireless communication service to the corresponding cell; Base Station-Cell Coverage, indicating that the base station's signal can cover the area where the cell is located; Root Cause-Cause-Reason, indicating that the root cause (e.g., base station antenna damage) leads to the direct cause (e.g., poor signal quality); Cause-Cause-Low Perception, indicating that the cause (e.g., poor signal quality) leads to a decrease in the user's perception of network service quality; Work Order-Occurring Cell-Cell, indicating that the complaint in the work order occurred within the corresponding cell.

[0099] In some embodiments, the server can train the above information extraction model in the following manner, which may specifically include: acquiring training data that identifies optimization schemes corresponding to different root causes, and using the training data to train a large language model to obtain the information extraction model.

[0100] This application does not limit the type of large language model in its embodiments. For example, such as... Figure 5 As shown, a large language model can be an LLaMA model, a chatGLM model, or a Baichuan model.

[0101] In the implementation of this application, the training data can be the same as the text information mentioned above, and there is no limitation here.

[0102] Specifically, taking the large language model as an example, specifically the LLaMA model, the server can convert the training data into a format that conforms to the requirements of the LLaMA model. Then, the server can label the converted training data, obtaining labeled training data. This labeled training data identifies the optimization schemes corresponding to different root causes. Then, as... Figure 6 As shown, the server can input training data, labeled with optimization schemes corresponding to different root causes, into the LLaMA model to train it, resulting in a trained LLaMA model. Then, the server can use human feedback reinforcement learning to optimize the trained LLaMA model, obtaining the final information extraction model.

[0103] The above methods can yield an information extraction model applicable to the field of network optimization, providing support for network optimization work.

[0104] In some embodiments, a root cause analysis model and a solution generation model are deployed on the server. The root cause analysis model is used to identify the target root cause that is causing the terminal's performance to be in a state of unoptimization, based on complaint tickets. The solution generation model is used to generate a target optimization solution based on the target root cause.

[0105] It should be noted that the solution generation model and the information extraction model can be separate models or integrated into a single model (hereinafter referred to as the network optimization model). The training method for the solution generation model can refer to the training method for the information extraction model described above, and will not be described here.

[0106] It should be noted that the root cause localization model is a lightweight traditional model. This application does not limit the type of root cause localization model. For example, as shown... Figure 7 As shown, the root cause localization model can be a graph neural network model, a convolutional neural network model, or a recurrent neural network model.

[0107] Specifically, taking the interaction between the root cause determination model and the network optimization model as an example, the above... Figure 2 S201-S203, specifically may include: such as Figure 8 As shown, server 102 can input complaint tickets into the network optimization model. The network optimization model can extract terminal performance-related information from the complaint tickets and generate a call instruction based on this information. This call instruction is then input into the root cause localization model, which instructs the root cause localization model to execute a target task. Subsequently, the root cause localization model executes the target task based on the call instruction, obtaining the target root cause output by the model that is causing the terminal's performance to be in a state of unoptimization. Finally, the root cause localization model can input the target root cause into the network optimization model to obtain the target optimization scheme output by the network optimization model.

[0108] This application does not limit the type of the target task in its embodiments. For example... Figure 8 As shown, the target tasks may include, but are not limited to: generating codes, retrieving databases, detecting network faults, analyzing the root causes of poor quality, generating text, and compressing alarms.

[0109] Figure 9 This is a schematic diagram of an optimization scheme determination device provided in an embodiment of this application, as shown below. Figure 9 As shown, the device includes:

[0110] The transceiver unit 901 is used to obtain complaint work orders sent by the terminal.

[0111] The determination unit 902 is used to determine the root cause that causes the terminal's performance to be in a state of unoptimization based on the complaint work order.

[0112] Processing unit 903 is used to generate target optimization schemes based on target root causes.

[0113] Optionally, the determining unit 902 is also used to determine the network device associated with the performance of the terminal.

[0114] Optionally, the processing unit 903 is also used to perform fault detection on the network device and obtain the detection results.

[0115] Optionally, the determining unit 902 is further configured to, when the detection results indicate that the network device has failed, determine the root cause of the failure and use the root cause as the target root cause that causes the terminal's performance to be in a state of unoptimization.

[0116] Optionally, the determining unit 902 is further configured to determine at least one optimization scheme corresponding to the target root cause.

[0117] Optionally, the determining unit 902 is further configured to determine a target optimization scheme based on at least one optimization scheme corresponding to the target root cause.

[0118] Optionally, the processing unit 903 is also used to extract the optimization scheme corresponding to each root cause from the textual information through an information extraction model.

[0119] Optionally, the processing unit 903 is also used to train the large language model using training data to obtain an information extraction model.

[0120] Figure 10 A schematic diagram of another possible structure of the optimization scheme determination device involved in the above embodiments is shown. This optimization scheme determination device includes a processor 1001 and a communication interface 1002. The processor 1001 is used to control and manage the operation of the optimization scheme determination device, and the communication interface 1002 is used to support communication between the optimization scheme determination device and other network entities. The optimization scheme determination device may also include a memory 1003 and a bus 1004, the memory 1003 being used to store the program code and data of the optimization scheme determination device.

[0121] The memory 1003 may be the memory in the optimization scheme determination device, etc. The memory may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as read-only memory, flash memory, hard disk or solid-state drive; the memory may also include a combination of the above types of memory.

[0122] The processor 1001 described above can implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with this disclosure. The processor can be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with this disclosure. The processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0123] Bus 1004 can be an extended industry standard architecture (EISA) bus, etc. Bus 1004 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 10 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0124] Through the above description of the embodiments, those skilled in the art will clearly understand that, for the sake of convenience and brevity, only the division of the above functional modules is used as an example. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device, and unit described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0125] This application provides a computer program product containing instructions that, when run on a computer, cause the computer to execute the optimization scheme determination method in the above method embodiments.

[0126] This application also provides a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the optimization scheme determination method in the method flow shown in the above method embodiments.

[0127] The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples (a non-exhaustive list) of computer-readable storage media include: electrical connections having one or more wires; portable computer disks; hard disks; random access memory (RAM); read-only memory (ROM); erasable programmable read-only memory (EPROM); registers; hard disks; optical fibers; compact disc read-only memory (CD-ROM); optical storage devices; magnetic storage devices; or any suitable combination thereof; or any other form of computer-readable storage medium known in the art. An exemplary storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the storage medium may also be a component of the processor. The processor and the storage medium may reside in an application-specific integrated circuit (ASIC). In the embodiments of this application, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0128] Embodiments of the present invention provide a computer program product containing instructions that, when executed on a computer, cause the computer to perform the optimization scheme determination method described in the embodiments of this application.

[0129] Since the optimization scheme determination device, computer-readable storage medium, and computer program product in the embodiments of the present invention can be applied to the above method, the technical effects obtained can also be referred to the above method embodiments, and the embodiments of the present invention will not be repeated here.

[0130] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0131] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0132] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0133] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for determining an optimization scheme, characterized in that, Applied to a server, the method comprises: obtaining a complaint work order sent by a terminal; the complaint work order is used to identify that the performance of the terminal is in a to-be-optimized state; based on the complaint work order, determining a target root cause that causes the performance of the terminal to be in a to-be-optimized state; the terminal is connected with a plurality of network devices, and the plurality of network devices include a target network device; the target root cause is a fault root cause of a fault of the target network device, and the target network device is a network device associated with the performance of the terminal; based on the target root cause and an optimization knowledge graph, generating a target optimization scheme; the target optimization scheme is used to improve the performance of the terminal; wherein the construction process of the optimization knowledge graph comprises: obtaining text information; the text information contains a plurality of root causes and the optimization scheme corresponding to each root cause in the plurality of root causes, and the plurality of root causes include the target root cause; through an information extraction model, the optimization scheme corresponding to each root cause is extracted from the text information; taking each root cause and the optimization scheme corresponding thereto as a first entity, a first knowledge graph is established with the first entity as a vertex and the corresponding relationship between the first entities as an edge; obtaining basic information of a network device related to each root cause in the plurality of root causes; taking each root cause and the basic information of the network device related thereto as a second entity, a second knowledge graph is established with the second entity as a vertex and the corresponding relationship between the second entities as an edge; by using entity alignment technology, the first knowledge graph and the second knowledge graph are fused to obtain the optimization knowledge graph.

2. The method of claim 1, wherein, The method further comprises: determining a network device associated with the performance of the terminal; performing fault detection on the network device to obtain a detection result; in the case that the detection result represents that the network device has a fault, determining a fault root cause of the fault of the network device; taking the fault root cause as the target root cause that causes the performance of the terminal to be in a to-be-optimized state.

3. The method of claim 2, wherein, The fault root cause represents that a first device of the network device has a fault; and the target optimization scheme is to repair the first device.

4. The method according to any one of claims 1 to 3, characterized in that, The method further comprises: from the optimization knowledge graph, determining at least one optimization scheme corresponding to the target root cause; based on the at least one optimization scheme corresponding to the target root cause, determining the target optimization scheme.

5. The method of claim 1, wherein, The method further comprises: the information extraction model is obtained by training in the following manner: obtaining training data identifying optimization schemes corresponding to different root causes; using the training data to train a large language model to obtain the information extraction model.

6. The method of claim 1, wherein, The server is deployed with a root cause determination model and a scheme generation model; the root cause determination model is used to determine a target root cause that causes the performance of the terminal to be in a to-be-optimized state based on the complaint work order; the scheme generation model is used to generate a target optimization scheme based on the target root cause.

7. An optimization scheme determination device, characterized in that, The device comprises: A transceiver unit is configured to acquire a complaint work order sent by a terminal, and the complaint work order is used to indicate that a performance of the terminal is in a to-be-optimized state. A determination unit is configured to determine, based on the complaint work order, a target root cause that causes the performance of the terminal to be in the to-be-optimized state, wherein the terminal is connected to a plurality of network devices, the plurality of network devices include a target network device, the target root cause is a fault root cause of a fault of the target network device, and the target network device is a network device associated with the performance of the terminal. A processing unit is configured to generate a target optimization scheme based on the target root cause and an optimization knowledge graph, and the target optimization scheme is used to improve the performance of the terminal. The construction process of the optimization knowledge graph includes: acquiring text information, wherein the text information includes the plurality of root causes and the optimization scheme corresponding to each root cause in the plurality of root causes, and the plurality of root causes include the target root cause; extracting, by an information extraction model, the optimization scheme corresponding to each root cause from the text information; taking each root cause and the optimization scheme corresponding to the root cause as a first entity, establishing a first knowledge graph with the first entity as a vertex and a corresponding relationship between the first entities as an edge; acquiring basic information of a network device related to each root cause in the plurality of root causes; taking each root cause and the basic information of the network device related to the root cause as a second entity, establishing a second knowledge graph with the second entity as a vertex and a corresponding relationship between the second entities as an edge; and fusing, by using an entity alignment technology, the first knowledge graph and the second knowledge graph to obtain the optimization knowledge graph.

8. An optimization scheme determination device, characterized in that, The method comprises the following steps: A processor and a communication interface are coupled, and the processor is configured to run a computer program or instructions to implement the optimization scheme determination method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores instructions, and when a computer executes the instructions, the computer executes the optimization scheme determination method according to any one of claims 1-6.

10. A computer program product, characterised in that, The computer program product contains computer instructions, and when the computer instructions are run on a computer, the computer instructions make the computer execute the optimization scheme determination method according to any one of claims 1-6.

Images