Fault processing method, device, equipment, readable storage medium and program product

Abstract

The application relates to a fault processing method, device, equipment, readable storage medium and program product. The method comprises the following steps: firstly, determining a fault scene of a target user in a network communication process according to state information of a network device of the target user; then, receiving fault information input by the target user, and performing intention recognition processing on the fault information according to the fault scene to obtain an intention recognition result; next, inputting the intention recognition result into a pre-trained fault elimination model, determining fault elimination information according to an output of the fault elimination model, and using the fault elimination information to indicate how to eliminate the network communication fault; finally, outputting the fault elimination information to the target user. The method can assist the target user in fault processing, and improves the efficiency of fault response processing in the network communication process.

Description

Technical Field

[0001] This application relates to the field of intelligent operation and maintenance technology, and in particular to a fault handling method, apparatus, device, readable storage medium, and program product. Background Technology

[0002] With the development of the internet, various operators have provided users with the core means to connect to the internet and obtain information. Home broadband installation and maintenance services ensure that users can connect to the internet. During the installation and maintenance of home broadband, faults may occur that require location and handling.

[0003] Currently, during the installation and maintenance of home broadband, when a fault occurs, it is impossible to provide specific solutions based on the specific cause of the fault. Summary of the Invention

[0004] Therefore, it is necessary to provide a fault handling method, apparatus, device, readable storage medium, and program product that can automatically handle faults based on input fault information, in order to address the above-mentioned technical problems.

[0005] Firstly, this application provides a fault handling method, including:

[0006] Determine the fault scenarios of the target user during network communication based on the status information of the target user's network devices;

[0007] Receive fault information input by the target user, and perform intent recognition processing on the fault information according to the fault scenario to obtain the intent recognition result;

[0008] The intent recognition results are input into a pre-trained troubleshooting model, and the troubleshooting information is determined based on the output of the troubleshooting model. The troubleshooting information is used to indicate how to troubleshoot network communication faults.

[0009] Output troubleshooting information to the target user.

[0010] In one embodiment, fault information input by a target user is received, and intent recognition processing is performed on the fault information according to the fault scenario to obtain intent recognition results, including:

[0011] The system receives multiple rounds of fault information from the target input. After each round of fault information input, the system performs intent recognition processing on the fault information input in this round according to the fault scenario to obtain intermediate intent recognition results.

[0012] If the fault information input in the previous round does not meet the fault troubleshooting conditions, then input prompt information is output based on the intermediate intent recognition result. The input prompt information is used to instruct the target user to input the fault information in the next round.

[0013] If the fault information input in the previous round meets the fault troubleshooting conditions, then the intent recognition result is obtained based on the fault information input by the user in multiple rounds.

[0014] In one embodiment, the fault information is processed for intent recognition based on the fault scenario to obtain the intent recognition result, including:

[0015] The fault scenario and fault information are input together into the intent recognition model;

[0016] The intent recognition result is obtained based on the output of the intent recognition model.

[0017] In one embodiment, inputting the intent recognition result into a pre-trained troubleshooting model includes:

[0018] The intent recognition result is encoded to obtain the encoded result, which includes absolute position encoding and relative position encoding.

[0019] The encoded processing results are input into the fault diagnosis model through the input projection layer, which is used to process multimodal data.

[0020] In one embodiment, determining fault troubleshooting information based on the output of the fault troubleshooting model includes:

[0021] The output of the fault diagnosis model is decoded after passing through the output projection layer, which is used to process multimodal data.

[0022] Based on the decoding results, troubleshooting information is determined.

[0023] In one embodiment, the troubleshooting model further includes a long short-term memory network for storing at least one of historical fault scenarios, historical fault information, and historical troubleshooting information.

[0024] In one embodiment, determining fault troubleshooting information based on the output of the fault troubleshooting model includes:

[0025] The fault troubleshooting information is determined based on the preset knowledge base, preset service interface, long short-term memory network, and fault troubleshooting model.

[0026] In one embodiment, before performing intent recognition processing on the fault information based on the fault scenario, the method further includes:

[0027] Perform a security verification on the fault information and determine the security verification result.

[0028] Secondly, this application also provides a fault handling apparatus, comprising:

[0029] The first determining module is used to determine the fault scenario of the target user in the network communication process based on the status information of the target user's network device;

[0030] The receiving module is used to receive fault information input by the target user, and perform intent recognition processing on the fault information according to the fault scenario to obtain the intent recognition result;

[0031] The second determining module is used to input the intent recognition result into the pre-trained fault diagnosis model, and determine the fault diagnosis information based on the output of the fault diagnosis model. The fault diagnosis information is used to indicate how to troubleshoot network communication faults.

[0032] The output module is used to output troubleshooting information to the target user.

[0033] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of any of the methods described in the first aspect above.

[0034] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described in the first aspect above.

[0035] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of any of the methods described in the first aspect above.

[0036] The aforementioned fault handling method, apparatus, device, readable storage medium, and program product first determine the fault scenario of the target user during network communication based on the status information of the target user's network device. Then, it receives fault information input by the target user and performs intent recognition processing on the fault information according to the fault scenario to obtain the intent recognition result. Next, it inputs the intent recognition result into a pre-trained fault troubleshooting model and determines fault troubleshooting information based on the output of the model. This fault troubleshooting information indicates how to troubleshoot the network communication fault. Finally, it outputs the fault troubleshooting information to the target user. In this way, upon receiving fault information input from the target user, the current fault scenario is determined, and fault troubleshooting information instructing on how to troubleshoot the network communication fault is obtained based on the fault scenario, fault information, and fault troubleshooting model. Finally, the fault troubleshooting information is displayed to the target user, assisting the target user in fault handling and improving the efficiency of fault response and handling during network communication. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is an application environment diagram of a fault handling method in one embodiment;

[0039] Figure 2 This is a flowchart illustrating a fault handling method in one embodiment;

[0040] Figure 3 This is a flowchart illustrating a fault handling method in another embodiment;

[0041] Figure 4 This is a flowchart illustrating a fault handling method in another embodiment;

[0042] Figure 5 This is a flowchart illustrating a fault handling method in another embodiment;

[0043] Figure 6 This is a schematic diagram of the relative position matrix in relative position encoding in one embodiment;

[0044] Figure 7 This is a schematic diagram showing the positions of the input projection layer and the output projection layer in another embodiment;

[0045] Figure 8 This is a flowchart illustrating a fault handling method in another embodiment;

[0046] Figure 9 This is a schematic diagram illustrating the organization of question-and-answer pair data in one embodiment;

[0047] Figure 10 Here is a flowchart of a fault handling method in another embodiment;

[0048] Figure 11 Here is a flowchart of a fault handling method in another embodiment;

[0049] Figure 12 This is a schematic diagram of the interface display of a fault handling method in one embodiment;

[0050] Figure 13 This is a schematic diagram of the structure of an Agent in one embodiment;

[0051] Figure 14 This is a structural block diagram of a fault handling device in one embodiment;

[0052] Figure 15 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0054] The fault handling method provided in this application embodiment can be applied to, for example, Figure 1 In the application environment shown, terminal 102 communicates with server 104 via a network. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. IoT devices can include smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, and projection devices. Portable wearable devices can include smartwatches, smart bracelets, and head-mounted displays. Head-mounted displays can be virtual reality (VR) devices, augmented reality (AR) devices, and smart glasses. Server 104 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0055] In one exemplary embodiment, such as Figure 2 As shown, a fault handling method is provided, which is applied to Figure 1 Taking the server in the example, the explanation includes the following steps 201 to 204. Wherein:

[0056] Step 201: Determine the fault scenario of the target user in the network communication process based on the status information of the target user's network devices.

[0057] Optionally, by integrating multiple sensors and data analysis technologies, the status information of the target user's network devices can be obtained from the home broadband network environment and physical space. This status information may include network speed, signal stability, device connection status, etc. Based on the status information, fault scenarios during the target user's home broadband network communication can be determined.

[0058] Step 202: Receive fault information input by the target user, and perform intent recognition processing on the fault information according to the fault scenario to obtain intent recognition results.

[0059] This process involves receiving requests initiated by the target user through their terminal, specifically receiving fault information input by the target user. Optionally, the fault information can be text messages such as "Why can't I access the internet with my home broadband?", or it can be image or video information, etc. The system stores, manages, and maintains the fault information, while also performing security verification to determine the results, thereby ensuring the accuracy and security of the fault information.

[0060] Optionally, after the security verification result is qualified, the fault information is processed for intent recognition according to the fault scenario. Intent recognition is used to understand the intent and goal of the target user in order to provide the target user with corresponding services or information. After intent recognition processing, the intent recognition result is obtained.

[0061] Step 203: Input the intent recognition result into the pre-trained troubleshooting model, and determine the troubleshooting information based on the output of the troubleshooting model.

[0062] The troubleshooting information is used to indicate how to troubleshoot network communication faults. Optionally, the intent recognition result is converted into a vector form and input into a pre-trained troubleshooting model, where the troubleshooting model can be a domain-wide model based on LLM (Large Language Model). Based on the inference of the troubleshooting model, the model processing result is output. Optionally, based on the model processing result, as well as other domain knowledge bases and service interfaces that need to be called, the troubleshooting information can be determined. The troubleshooting information can indicate how to troubleshoot the fault corresponding to the fault information input by the target user.

[0063] Step 204: Output troubleshooting information to the target user.

[0064] Troubleshooting information is output to the target user's terminal for troubleshooting. Optionally, the troubleshooting information can be structured and arranged in an orderly manner to ensure that it is displayed on the terminal in a logical and coherent way. Optionally, the troubleshooting information can be displayed in various ways, such as text, images, or videos.

[0065] In the above embodiments, firstly, the fault scenario of the target user during network communication is determined based on the status information of the target user's network device. Then, fault information input by the target user is received, and intent recognition processing is performed on the fault information according to the fault scenario to obtain the intent recognition result. Next, the intent recognition result is input into a pre-trained fault troubleshooting model, and fault troubleshooting information is determined based on the output of the fault troubleshooting model. The fault troubleshooting information is used to indicate how to troubleshoot network communication faults. Finally, the fault troubleshooting information is output to the target user. In this way, after receiving fault information input by the target user, the current fault scenario is determined, and fault troubleshooting information indicating how to troubleshoot network communication faults can be obtained based on the fault scenario, fault information, and fault troubleshooting model. Finally, the fault troubleshooting information is displayed to the target user to assist the target user in troubleshooting, thereby improving the efficiency of fault response and handling during network communication.

[0066] In the embodiments of this application, step 202 described above can be as follows: Figure 3 As shown, it includes:

[0067] Step 301: Receive multiple rounds of fault information input by the target user. After each round of fault information input, perform intent recognition processing on the fault information input in this round according to the fault scenario to obtain intermediate intent recognition results.

[0068] Optionally, in order to better identify and confirm the true intent of the target user and provide more accurate services or responses, multiple rounds of fault information from the target user can be received through multi-round dialogue, and intent recognition processing can be performed on the fault information input in each round.

[0069] Step 302: If the fault information input in the previous round does not meet the fault troubleshooting conditions, then output input prompt information based on the intermediate intent recognition result.

[0070] The input prompt information is used to instruct the target user to input the next round of fault information. If the fault information input by the target user in the previous round does not meet the fault troubleshooting conditions, that is, the content of the fault information is insufficient and cannot be used for fault troubleshooting, the target user needs to input more auxiliary information. Therefore, based on the intermediate intent recognition result obtained from the fault information input in the previous round, the input prompt information is output to instruct the target user to input the next round of fault information, so as to increase the effective fault information for fault troubleshooting.

[0071] Step 303: If the fault information input in the previous round meets the fault elimination conditions, then the intent recognition result is obtained based on the fault information input by the target user in multiple rounds.

[0072] If the fault information input by the target user in the previous rounds has met the fault troubleshooting conditions, then the intent recognition process is performed based on the fault information input by the target user in all previous rounds to obtain the intent recognition result.

[0073] In the above embodiments, during the interaction with the target user, multi-turn dialogue is used to better identify and confirm the target user's true intention, thereby improving the accuracy of fault handling for the target user and enhancing the user's service experience.

[0074] Optionally, the steps for obtaining the intent recognition result through intent recognition processing can be as follows: Figure 4 As shown, it includes:

[0075] Step 401: Input the fault scenario and fault information into the intent recognition model.

[0076] The fault information and fault scenario input by the target user in multi-turn dialogue are processed by the intent recognition model. Optionally, the intent recognition model can be a model based on natural language processing technology or a model based on deep learning.

[0077] Step 402: Obtain the intent recognition result based on the output of the intent recognition model.

[0078] Based on the output of the intent recognition model, the intent recognition result is determined. Then, the intent recognition result is converted into vector form and input into the troubleshooting model for inference.

[0079] In one embodiment, to achieve multimodal data fusion, the step of inputting the intent recognition result into a pre-trained fault diagnosis model is as follows: Figure 5 As shown, it includes:

[0080] Step 501: Encode the intent recognition result to obtain the encoded result.

[0081] The encoding process includes absolute positional encoding and relative positional encoding. Currently, the tokens input into the LLM are mostly unordered data. To preserve the positional information of the input vector, positional information needs to be added. This is typically achieved using absolute positional embedding, where each element in the vector represents a specific position in the sentence. The element embedding and positional embedding are simply added together as the input to the LLM. Relative positional encoding, on the other hand, does not represent the absolute position of a token in the sentence; it learns the representation of each pair of tokens in the sentence. Since the positions of phrases are all different, this application improves the attention mechanism by incorporating relative positional embeddings.

[0082] like Figure 6 As shown, , … This represents the relative distance between each character, and the relative distance is added to the query-key matrix product of the self-attention layer. The rotation matrix... It is to rotate the vector angle, It is the absolute position of the token in the sentence. This is the vector being rotated (taking a two-dimensional vector as an example). Before using the rotation matrix, a linear transformation is used to obtain... and The value of is determined to preserve rotational distortion, as shown in the following formula:

[0083] (1)

[0084] Step 502: Input the encoding processing result into the fault diagnosis model through the input projection layer.

[0085] The input projection layer is used to process multimodal data. After encoding, the encoded results are passed through the input projection layer and then input into the LLM, such as... Figure 7 As shown, the InputProjection layer, which is trained between the LLM and the encoding layer, serves as a mapping relationship between multimodal data such as images, audio, and video, etc., and the LLM.

[0086] In one embodiment, the step of determining fault troubleshooting information based on the output of the fault troubleshooting model is as follows: Figure 8 As shown, it includes:

[0087] Step 801: The output of the fault diagnosis model is decoded after passing through the output projection layer.

[0088] The output projection layer is used to process multimodal data. Please refer to [link / reference needed]. Figure 7 The LLM and decoding layers are trained to form an output projection layer, which serves as a mapping between the LLM output and multimodal data such as images, audio, and video. Optionally, additional knowledge bases can be added to the LLM to enable fine-tuning of instructions, allowing the LLM to learn and transform between multimodal data.

[0089] Step 802: Determine the troubleshooting information based on the decoding results.

[0090] Based on the results of the decoding process, troubleshooting information is obtained, and then the troubleshooting information is output to the target user.

[0091] Optionally, to improve the accuracy of the fault troubleshooting model's inference, the fault troubleshooting model also includes a long short-term memory network, which is used to store at least one of historical fault scenarios, historical fault information, and historical fault troubleshooting information.

[0092] Regarding long-term memory, the long-term memory mechanism of the LLM model can be divided into two main categories. The first type of memory method involves embedding memory modules into the LLM model, i.e., the knowledge reserves within the neural network parameter system. This type of memory originates from the model's pre-learning and fine-tuning process of information in the communication domain, endowing the LLM model with basic cognitive and memory capabilities. For the intelligent agent in home broadband installation and maintenance proposed in this application, fine-tuning can be performed by introducing existing knowledge bases specific to home broadband installation and maintenance question-and-answer scenarios, thereby enhancing the understanding and response capabilities to fault information of the target user. Optionally, the data source question-and-answer pairs can be as follows: Figure 9 As shown. Another type of memory function can be indirectly achieved through external auxiliary tools and mechanisms. This method allows the LLM model to access external resources, such as external databases, knowledge graphs, or specialized search engines, when needed to supplement or enhance its own memory capabilities, enabling real-time access and utilization of broader and more dynamic information. This memory method not only expands the memory boundaries of the LLM model but also improves its adaptability and flexibility in complex and changing environments.

[0093] In large LLM models, short-term memory is mainly used for the context window of the large model, and all context learning is done by utilizing the model's short-term memory.

[0094] As shown above, long-term and short-term memories in the Long Short-Term Memory (LLM) network interact and cooperate within the large LLM model. Short-term memory provides new information and cues to long-term memory, promoting the formation and updating of long-term memory. Long-term memory can provide necessary background knowledge and experience support for short-term memory in processing fault information. Optionally, this can include at least one of historical fault scenarios, historical fault information, and historical fault troubleshooting information, thereby improving the processing efficiency and accuracy of short-term memory.

[0095] In the above embodiments, by adding a long short-term memory network, the interaction and collaboration mechanism of the long short-term memory network enables the LLM large model to maintain continuity and efficiency when processing complex tasks, while also supporting the continuous learning and evolution of the LLM large model.

[0096] Optionally, troubleshooting information can be determined based on a preset knowledge base, preset service interfaces, long short-term memory networks, and troubleshooting models.

[0097] The preset knowledge base can provide necessary domain knowledge for home broadband installation and maintenance, and the preset service interface can be provided by other service providers. These services can be used to assist the LLM model in reasoning, thereby improving the accuracy and professionalism of the LLM model inference and providing more accurate fault handling solutions for target users.

[0098] In embodiments of this application, a fault handling method is provided, which can be applied to, for example... Figure 10 The service process shown analyzes the current state of the home broadband installation and maintenance field, revealing insufficient convenience of self-service. This fault handling method, through the LLM-Agent model, enables self-learning, self-planning, and self-execution. Specifically, expert experience replication integrates existing mature diagnostic and repair processes into the intelligent agent; task self-learning continuously learns expert orchestration logic and massive user fault handling trajectories to achieve self-learning of fault handling tasks; task self-planning accurately identifies fault points based on user issues and intelligently plans the API call flow for diagnostic or repair capabilities; and task self-execution, based on the planned process, uses a large model to extract detection parameters from the dialogue context for competitive detection, rapid repair, and outputs operation manuals or operation videos based on user actions.

[0099] The fault handling method may include, for example: Figure 11 The modules shown illustrate how this fault handling method, based on an LLM-Agent (intelligent agent), provides interactive troubleshooting, acquires the target user's needs, and autonomously formulates the optimal fault handling solution. The target user inputs fault information through the Agent entry point. The fault handling method's process includes the following steps:

[0100] ①Target users can initiate a request to input fault information and interact with the LLM inference module through clients, including computers and mobile phones.

[0101] ② The content security module receives fault information and manages and maintains the collected fault information during the target user's intent recognition and processing to ensure the accuracy and security of the fault information.

[0102] ③ Obtain fault information through a multi-turn dialogue module.

[0103] ④ The intent recognition module performs intent recognition processing on the target user's fault information. The intent recognition module includes an intent recognition model.

[0104] ⑤ Obtain fault scenario information through the Agent module. Optionally, in home broadband installation and maintenance scenarios, LLM-Agent activates its powerful perception capabilities, integrating multiple sensors and data analysis technologies to collect fault scenario information from the home broadband network environment and physical space, including but not limited to potential fault points such as network speed, signal stability, and device connection status.

[0105] ⑥ Use intelligent orchestration technology to manage and optimize information processing workflows, while utilizing memory modules to store and process key data for use in subsequent steps.

[0106] ⑦ Convert the intent recognition results into vector form and input them into the domain-wide model for inference.

[0107] ⑧ The memory module stores at least one of the following: historical fault scenarios, historical fault information, and historical fault troubleshooting information, along with other relevant information, for use in subsequent interactions. Capability Provider A and Capability Provider B can provide specific functions or services to meet the needs of the target user. Various available service interfaces and APIs are registered through DCOOS. The API and scenario information management module can synchronize information across all modules, ensuring that all relevant components have the latest data. When an API call is made, the Agent uses the functions or services obtained from the capability providers to complete the corresponding task.

[0108] ⑨ During the Agent capability execution phase, the Agent outputs information based on the troubleshooting information obtained from the LLM large model inference module.

[0109] ⑩ Domain-wide models and capability providers can provide the necessary domain knowledge and capability support for determining troubleshooting information, ensuring the accuracy and professionalism of the output information.

[0110] ⑪ The process of structuring and organizing troubleshooting information or content through content arrangement ensures that the information is presented to the target user in a reasonable, coherent, and engaging manner. Optionally, the entire troubleshooting process can be as follows: Figure 12 As shown, the existing complex button-based interaction mode is changed to a simple dialogue interaction that enables accurate detection of fault points and shortens the detection time.

[0111] Optional, such as Figure 13 As shown, the Agent, through a cyclical logic of "perception -> planning -> action," can efficiently handle home broadband installation and maintenance tasks. Perception involves acquiring fault environment information. Planning, through the Agent's built-in algorithms and policy library, develops a detailed installation and maintenance solution based on the current network status and user needs. After planning is complete, the LLM-Agent enters the capability execution phase. According to the plan, it automatically or guides manual operations such as network equipment debugging, line rearrangement, and software upgrades. During action, the Agent continuously observes environmental changes and target user feedback, obtains fault information input by the target user, and adjusts action strategies in real time. During action, it can call external APIs to obtain additional information not included in the LLM large model weights, including real-time information, code execution capabilities, and access to dedicated information sources.

[0112] In the above embodiments, the LLM-Agent-based fault handling method provided in this application can offer interactive troubleshooting during home broadband installation and maintenance. It can understand the needs of target users, independently formulate the optimal detection plan, and output fault resolution information. Target users can access the solution directly with one click through the Agent entry embedded in the homepage. Combined with intent awareness and precise recommendations of image and text repair solutions, it improves the self-closing rate of faults during network communication.

[0113] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0114] Based on the same inventive concept, this application also provides a fault handling apparatus for implementing the fault handling method described above. The solution provided by this apparatus is similar to the solution described in the above method; therefore, the specific limitations in one or more fault handling apparatus embodiments provided below can be found in the limitations of the fault handling method described above, and will not be repeated here.

[0115] In one exemplary embodiment, such as Figure 14 As shown, a fault handling device 1400 is provided, comprising: a first determining module, a receiving module, a second determining module, and an output module, wherein:

[0116] The first determining module is used to determine the fault scenario of the target user in the network communication process based on the status information of the target user's network device;

[0117] The receiving module is used to receive fault information input by the target user, and perform intent recognition processing on the fault information according to the fault scenario to obtain the intent recognition result;

[0118] The second determining module is used to input the intent recognition result into the pre-trained fault diagnosis model, and determine the fault diagnosis information based on the output of the fault diagnosis model. The fault diagnosis information is used to indicate how to troubleshoot network communication faults.

[0119] The output module is used to output troubleshooting information to the target user.

[0120] In one embodiment, the receiving module is specifically used to receive multiple rounds of fault information input by the target. After each round of fault information input, the fault information input in this round is processed for intent recognition based on the fault scenario to obtain an intermediate intent recognition result. If the fault information input in the previous round does not meet the fault elimination conditions, input prompt information is output based on the intermediate intent recognition result. The input prompt information is used to instruct the target user to input the next round of fault information. If the fault information input in the previous round meets the fault elimination conditions, an intent recognition result is obtained based on the multiple rounds of fault information input by the user.

[0121] In one embodiment, the receiving module is specifically used to input the fault scenario and fault information into the intent recognition model; and obtain the intent recognition result based on the output of the intent recognition model.

[0122] In one embodiment, the second determining module is specifically used to encode the intent recognition result to obtain an encoding result, the encoding process including absolute position encoding and relative position encoding; and to input the encoding result into the fault diagnosis model through the input projection layer, the input projection layer being used to process multimodal data.

[0123] In one embodiment, the second determining module is specifically used to decode the output result of the fault troubleshooting model after passing it through the output projection layer, the output projection layer being used to process multimodal data; and to determine fault troubleshooting information based on the result of the decoding process.

[0124] In one embodiment, the troubleshooting model further includes a long short-term memory network for storing at least one of historical fault scenarios, historical fault information, and historical troubleshooting information.

[0125] In one embodiment, the second determining module is specifically used to determine fault troubleshooting information based on a preset knowledge base, a preset service interface, a long short-term memory network, and a fault troubleshooting model.

[0126] In one embodiment, the device further includes a security verification module for performing security verification on fault information and determining the security verification result.

[0127] Each module in the aforementioned fault handling device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0128] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 15 As shown, this computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores fault information data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When the computer program is executed by the processor, it implements a fault handling method.

[0129] Those skilled in the art will understand that Figure 15 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0130] In one exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: determining a fault scenario of the target user during network communication based on the status information of the target user's network device; receiving fault information input by the target user and performing intent recognition processing on the fault information according to the fault scenario to obtain an intent recognition result; inputting the intent recognition result into a pre-trained fault troubleshooting model, determining fault troubleshooting information based on the output of the fault troubleshooting model, the fault troubleshooting information being used to indicate how to troubleshoot network communication faults; and outputting the fault troubleshooting information to the target user.

[0131] In one embodiment, when the processor executes the computer program, it further implements the following steps: receiving multiple rounds of fault information input by the target, wherein, after each round of fault information input, the fault information input in this round is subjected to intent recognition processing according to the fault scenario to obtain an intermediate intent recognition result; if the fault information input in the previous round does not meet the fault elimination conditions, input prompt information is output according to the intermediate intent recognition result, and the input prompt information is used to instruct the target user to input the next round of fault information; if the fault information input in the previous round meets the fault elimination conditions, an intent recognition result is obtained according to the multiple rounds of fault information input by the user.

[0132] In one embodiment, when the processor executes the computer program, it further performs the following steps: inputting the fault scenario and fault information into the intent recognition model; and obtaining the intent recognition result based on the output of the intent recognition model.

[0133] In one embodiment, when the processor executes the computer program, it further performs the following steps: encoding the intent recognition result to obtain an encoding result, the encoding process including absolute position encoding and relative position encoding; inputting the encoding result into a fault diagnosis model through an input projection layer, the input projection layer being used to process multimodal data.

[0134] In one embodiment, when the processor executes the computer program, it further performs the following steps: decoding the output of the fault troubleshooting model after passing it through an output projection layer, wherein the output projection layer is used to process multimodal data; and determining fault troubleshooting information based on the result of the decoding process.

[0135] In one embodiment, the troubleshooting model further includes a long short-term memory network for storing at least one of historical fault scenarios, historical fault information, and historical troubleshooting information.

[0136] In one embodiment, when the processor executes the computer program, it also performs the following steps: determining troubleshooting information based on a preset knowledge base, a preset service interface, a long short-term memory network, and a troubleshooting model.

[0137] In one embodiment, when the processor executes the computer program, it also performs the following steps: performing a security verification on the fault information and determining the security verification result.

[0138] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon. When executed by a processor, the computer program performs the following steps: determining a fault scenario of the target user during network communication based on the status information of the target user's network device; receiving fault information input by the target user and performing intent recognition processing on the fault information according to the fault scenario to obtain an intent recognition result; inputting the intent recognition result into a pre-trained fault troubleshooting model, determining fault troubleshooting information based on the output of the fault troubleshooting model, the fault troubleshooting information being used to indicate how to troubleshoot network communication faults; and outputting the fault troubleshooting information to the target user.

[0139] In one embodiment, when the computer program is executed by the processor, it further implements the following steps: receiving multiple rounds of fault information input from the target, wherein, after each round of fault information input, the fault information input in this round is subjected to intent recognition processing according to the fault scenario to obtain an intermediate intent recognition result; if the fault information input in the previous round does not meet the fault elimination conditions, input prompt information is output according to the intermediate intent recognition result, and the input prompt information is used to instruct the target user to input the next round of fault information; if the fault information input in the previous round meets the fault elimination conditions, an intent recognition result is obtained according to the multiple rounds of fault information input by the user.

[0140] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: inputting the fault scenario and fault information into the intent recognition model; and obtaining the intent recognition result based on the output of the intent recognition model.

[0141] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: encoding the intent recognition result to obtain an encoding result, the encoding process including absolute position encoding and relative position encoding; inputting the encoding result into a fault diagnosis model through an input projection layer, the input projection layer being used to process multimodal data.

[0142] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: decoding the output of the fault diagnosis model after passing it through an output projection layer, wherein the output projection layer is used to process multimodal data; and determining fault diagnosis information based on the result of the decoding process.

[0143] In one embodiment, the troubleshooting model further includes a long short-term memory network for storing at least one of historical fault scenarios, historical fault information, and historical troubleshooting information.

[0144] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: determining troubleshooting information based on a preset knowledge base, a preset service interface, a long short-term memory network, and a troubleshooting model.

[0145] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: performing a security verification on the fault information and determining the security verification result.

[0146] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps: determining a fault scenario of the target user during network communication based on the status information of the target user's network device; receiving fault information input by the target user and performing intent recognition processing on the fault information according to the fault scenario to obtain an intent recognition result; inputting the intent recognition result into a pre-trained fault troubleshooting model, determining fault troubleshooting information based on the output of the fault troubleshooting model, the fault troubleshooting information being used to indicate how to troubleshoot network communication faults; and outputting the fault troubleshooting information to the target user.

[0147] In one embodiment, when the computer program is executed by the processor, it further implements the following steps: receiving multiple rounds of fault information input from the target, wherein, after each round of fault information input, the fault information input in this round is subjected to intent recognition processing according to the fault scenario to obtain an intermediate intent recognition result; if the fault information input in the previous round does not meet the fault elimination conditions, input prompt information is output according to the intermediate intent recognition result, and the input prompt information is used to instruct the target user to input the next round of fault information; if the fault information input in the previous round meets the fault elimination conditions, an intent recognition result is obtained according to the multiple rounds of fault information input by the user.

[0148] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: inputting the fault scenario and fault information into the intent recognition model; and obtaining the intent recognition result based on the output of the intent recognition model.

[0149] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: encoding the intent recognition result to obtain an encoding result, the encoding process including absolute position encoding and relative position encoding; inputting the encoding result into a fault diagnosis model through an input projection layer, the input projection layer being used to process multimodal data.

[0150] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: decoding the output of the fault diagnosis model after passing it through an output projection layer, wherein the output projection layer is used to process multimodal data; and determining fault diagnosis information based on the result of the decoding process.

[0151] In one embodiment, the troubleshooting model further includes a long short-term memory network for storing at least one of historical fault scenarios, historical fault information, and historical troubleshooting information.

[0152] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: determining troubleshooting information based on a preset knowledge base, a preset service interface, a long short-term memory network, and a troubleshooting model.

[0153] In one embodiment, when the computer program is executed by the processor, it further performs the following steps: performing a security verification on the fault information and determining the security verification result.

[0154] 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, data stored, data displayed, 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 the relevant data must comply with relevant regulations.

[0155] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0156] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0157] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A fault handling method, characterized in that, The method includes: The fault scenario of the target user during network communication is determined based on the status information of the target user's network device; Receive fault information input by the target user, and perform intent recognition processing on the fault information according to the fault scenario to obtain intent recognition result; The intent recognition result is input into a pre-trained troubleshooting model, and troubleshooting information is determined based on the output of the troubleshooting model. The troubleshooting information is used to indicate how to troubleshoot network communication faults. The troubleshooting information is output to the target user.

2. The method according to claim 1, characterized in that, The process of receiving fault information input by the target user and performing intent recognition processing on the fault information according to the fault scenario to obtain intent recognition results includes: The system receives multiple rounds of fault information input from the target user. After each round of fault information input, the system performs intent recognition processing on the fault information input in this round according to the fault scenario to obtain an intermediate intent recognition result. If the fault information input in the previous round does not meet the fault troubleshooting conditions, then input prompt information is output according to the intermediate intent recognition result. The input prompt information is used to instruct the target user to input the fault information in the next round. If the fault information input in the previous round meets the fault elimination conditions, then the intent recognition result is obtained based on the fault information input by the target user in multiple rounds.

3. The method according to claim 1, characterized in that, The step of performing intent recognition processing on the fault information based on the fault scenario to obtain the intent recognition result includes: The fault scenario and the fault information are input together into the intent recognition model; The intent recognition result is obtained based on the output of the intent recognition model.

4. The method according to claim 1, characterized in that, The step of inputting the intent recognition result into a pre-trained fault diagnosis model includes: The intent recognition result is encoded to obtain an encoding result, wherein the encoding process includes absolute position encoding and relative position encoding. The encoding processing result is input into the fault diagnosis model through the input projection layer, which is used to process multimodal data.

5. The method according to claim 4, characterized in that, Determining fault troubleshooting information based on the output of the fault troubleshooting model includes: The output of the fault diagnosis model is decoded after passing through the output projection layer, which is used to process multimodal data. Based on the result of the decoding process, the troubleshooting information is determined.

6. The method according to claim 5, characterized in that, The troubleshooting model also includes a long short-term memory network, which is used to store at least one of historical fault scenarios, historical fault information, and historical troubleshooting information.

7. The method according to claim 6, characterized in that, Determining fault troubleshooting information based on the output of the fault troubleshooting model includes: The fault troubleshooting information is determined based on the preset knowledge base, the preset service interface, the long short-term memory network, and the fault troubleshooting model.

8. The method according to claim 1, characterized in that, The method further includes performing intent recognition processing on the fault information based on the fault scenario: The fault information is subjected to a security verification, and the security verification result is determined.

9. A fault handling device, characterized in that, The device includes: The first determining module is used to determine the fault scenario of the target user in the network communication process based on the status information of the target user's network device; The receiving module is used to receive fault information input by the target user, and perform intent recognition processing on the fault information according to the fault scenario to obtain intent recognition results; The second determining module is used to input the intent recognition result into a pre-trained troubleshooting model, and determine troubleshooting information based on the output of the troubleshooting model. The troubleshooting information is used to indicate how to troubleshoot network communication faults. The output module is used to output the troubleshooting information to the target user.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 8.

Images