Problem solution obtaining method and device based on large model and electronic equipment

Abstract

The invention provides a problem solution obtaining method and device based on a large model and electronic equipment, and relates to the technical field of artificial intelligence, in particular to the application fields of generative search, intelligent document editing, intelligent assistants, virtual humans and the like. According to the specific implementation scheme, current error reporting information of a target system is obtained; acquiring prompt information; wherein the prompt information comprises N groups of optimization reference information; n is an integer greater than or equal to 1; the optimization reference information comprises historical error reporting information related to the target system and a historical problem solution for the historical error reporting information; and obtaining a current problem solution for the current error report information by using the large model based on the prompt information.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and in particular to application fields such as generative search, intelligent document editing, intelligent assistants, and virtual humans, and specifically to a method, device, and electronic device for obtaining a problem solution based on a large model. Background Art

[0002] As a comprehensive system built according to specific information needs, the information retrieval system integrates programmed functions such as information collection, fine processing, efficient storage and intelligent retrieval. However, given the inherent complexity and technical depth of the information retrieval system, when it encounters system error information, it is often necessary to locate the problem through manual intervention in order to develop a solution to the problem of the system error information. This process is not only inefficient, but also significantly increases the burden on human resources, posing a serious challenge to the timeliness and cost-effectiveness of the operation and maintenance of the information retrieval system. Summary of the invention

[0003] The present disclosure provides a method, device and electronic device for obtaining a problem solution based on a large model.

[0004] According to a first aspect of the present disclosure, a method for obtaining a problem solution based on a large model is provided, comprising:

[0005] Get the current error information of the target system;

[0006] Obtain prompt information; wherein the prompt information includes N groups of optimization reference information; N ≥ 1 and is an integer; the optimization reference information includes historical error information related to the target system and historical problem solutions for the historical error information;

[0007] Using the big model, based on the prompt information, we can get the current problem solution for the current error message.

[0008] According to a second aspect of the present disclosure, there is provided a device for obtaining a problem solution based on a large model, comprising:

[0009] An error information acquisition unit, used to acquire current error information of the target system;

[0010] A prompt information acquisition unit, used to acquire prompt information; wherein the prompt information includes N groups of optimization reference information; N ≥ 1 and is an integer; the optimization reference information includes historical error information related to the target system and historical problem solutions for the historical error information;

[0011] The problem solution obtaining unit is used to obtain the current problem solution for the current error information based on the prompt information by using the large model.

[0012] According to a third aspect of the present disclosure, there is provided an electronic device, including:

[0013] at least one processor;

[0014] a memory communicatively coupled to the at least one processor;

[0015] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method provided in the first aspect of the present disclosure.

[0016] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to cause the computer to execute the method provided according to the first aspect of the present disclosure.

[0017] According to a fifth aspect of the present disclosure, a computer program product is provided, comprising a computer program, which implements the method provided according to the first aspect of the present disclosure when executed by a processor.

[0018] The use of the present disclosure can improve the timeliness of operation and maintenance of the target system and reduce the maintenance cost of the target system.

[0019] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0021] Figure 1 A flowchart of a method for obtaining a problem solution based on a large model provided in an embodiment of the present disclosure;

[0022] Figure 2 A flowchart of a method for obtaining a problem solution based on a large model provided in an embodiment of the present disclosure;

[0023] Figure 3 An auxiliary illustration of the completeness process of a method for obtaining a problem solution based on a large model provided in an embodiment of the present disclosure;

[0024] Figure 4 A schematic diagram of a scenario of a method for obtaining a problem solution based on a large model provided in an embodiment of the present disclosure;

[0025] Figure 5 A schematic structural block diagram of a device for obtaining a problem solution based on a large model provided in an embodiment of the present disclosure;

[0026] Figure 6 A schematic structural block diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0027] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present disclosure. Similarly, for the sake of clarity and conciseness, the description of well-known functions and structures is omitted in the following description.

[0028] The embodiment of the present disclosure provides a method for obtaining a problem solution based on a large model, which can be applied to an electronic device. The electronic device can be a server, a workbench, a mainframe computer, a conventional computer (desktop computer, laptop computer, etc.) or other similar computing devices. Figure 1 The flowchart shown in the figure illustrates a method for obtaining a problem solution based on a large model provided by an embodiment of the present disclosure. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described in the flowchart may also be performed in other orders.

[0029] Step S101, obtaining current error information of the target system.

[0030] The target system may be an information retrieval system. Here, the information retrieval system may be an overall retrieval system, such as a search engine, a document retrieval system, a multimedia retrieval system, a database retrieval system, an academic retrieval system, etc., or a subsystem in the overall retrieval system, such as a retrieval subsystem, a database creation and maintenance subsystem, a sorting subsystem, etc.

[0031] In addition, it is understood that in the embodiments of the present disclosure, when an abnormal situation occurs in the target system, a current error message will be generated and issued. The abnormal situation may include the inability to find the search resource, internal service error, request rejection, request format error, too many requests, network problem, security problem, etc.

[0032] Step S102, obtaining prompt information.

[0033] The prompt information includes N groups of optimization reference information, where N≥1 and is an integer; the optimization reference information includes historical error information related to the target system and historical problem solutions for the historical error information.

[0034] In addition, it should be noted that in the embodiments of the present disclosure, historical problem solutions can be obtained through manual intervention, or can be derived from the system log of the target system, or can be derived from the knowledge base of the target system. The knowledge base of the target system is used to store the architecture information, component information, configuration information, function information, interface documents, data models, business specifications, operation procedures, common problems and their solutions, etc. of the target system.

[0035] Step S103, using the large model, based on the prompt information, obtain a current problem solution for the current error information.

[0036] The large model may be a large language model, specifically a pre-trained neural network model, which has general language knowledge, world knowledge, professional knowledge in various fields, etc. In one example, the large model may be an autoregressive generative model of a Transformer architecture.

[0037] After obtaining the current error information and prompt information of the target system, the current error information and prompt information can be input into the big model to use the big model to obtain the current problem solution for the current error information based on N groups of optimization reference information included in the prompt information.

[0038] By adopting the problem-solving method acquisition method based on a large model provided by the embodiment of the present disclosure, it is possible to acquire the current error information of the target system, as well as acquire prompt information (wherein the prompt information includes N groups of optimization reference information; N ≥ 1 and is an integer; the optimization reference information includes historical error information related to the target system, and historical problem solutions for the historical error information), and utilize the large model to obtain the current problem solution for the current error information based on the prompt information. In this way, compared with the prior art, which requires manual intervention to locate the problem in order to formulate a problem-solving solution for the system error information, it can not only improve the efficiency of solving system problems, but also significantly reduce the burden on human resources, and ultimately achieve the purpose of the invention of improving the timeliness of operation and maintenance of the target system and reducing the maintenance cost of the target system.

[0039] In some optional implementations, step S102, i.e., “obtaining prompt information” may include:

[0040] Based on the system log of the target system, M groups of first reference information are obtained; wherein M≥1 and is an integer; the first reference information includes first error information related to the target system and a first problem solution for the first error information;

[0041] Based on the knowledge base of the target system, Z groups of second reference information are obtained; wherein Z ≥ 1 and is an integer; the second reference information includes second error information related to the target system and a second problem solution for the second error information;

[0042] Based on the M groups of first reference information and the Z groups of second reference information, N groups of optimized reference information are obtained;

[0043] Based on N sets of optimization reference information, prompt information is obtained.

[0044] In one example, “obtaining M sets of first reference information based on the system log of the target system” may include:

[0045] Determine multiple candidate error messages from the system log of the target system;

[0046] Selecting M first error messages from a plurality of candidate error messages; wherein the first error messages are candidate error messages for which corresponding first problem solutions exist in the system log of the target system;

[0047] M groups of first reference information are obtained by using the M pieces of first error information and the M first problem solutions corresponding one-to-one to the M pieces of first error information.

[0048] In a specific example, multiple candidate error messages can be determined from the system log of the target system by keyword retrieval. In this process, the keywords involved may include error code, error description, exception type, etc. Here, error code may include "500", "404", etc.; error description may include "Error", "Critical", "Exception", "Failed", etc.; exception type may include null pointer exception (also known as Null PointerException), memory overflow error (also known as Out Of Memory Error), etc.

[0049] In addition, in the above example, after selecting M first error messages from multiple candidate error messages, M first problem solutions corresponding to the M first error messages can also be determined from the system log of the target system. For example, it is also possible to determine the M first problem solutions corresponding to the M first error messages from the system log of the target system by keyword retrieval. In this process, the keywords involved may include repair steps, patches or updates, configuration changes, etc. Among them, repair steps may include "Reinstall", "Update driver", "RestartService", etc.; patches or updates may include "Patch 1.0.1", "Update To Version 2.2", etc.; configuration changes may include "Modify Configuration File", "Change Settings", etc.

[0050] After obtaining M first error messages and M first problem solutions corresponding one-to-one to the M first error messages, the first error messages and first problem solutions that have a corresponding relationship among the M first error messages and the M first problem solutions can be combined into a group of first reference information, and finally, M groups of first reference information are obtained.

[0051] In the above example, multiple candidate error messages can be automatically determined from the system log of the target system, and M first error messages are selected from the multiple candidate error messages (the first error messages are candidate error messages for which corresponding first problem solutions exist in the system log of the target system), and then M sets of first reference information are obtained by using the M first error messages and the M first problem solutions corresponding to the M first error messages. In this way, the efficiency of obtaining the M sets of first reference information can be improved, so as to further improve the efficiency of solving system problems, thereby further improving the timeliness of operation and maintenance of the target system.

[0052] In one example, “based on the knowledge base of the target system, obtain Z sets of second reference information, including:

[0053] Selecting Z second error messages from the plurality of remaining error messages; wherein the plurality of remaining error messages are other candidate error messages except the M first error messages from the plurality of candidate error messages; the second error messages are error messages for which a corresponding second problem solution can be retrieved in a knowledge base of the target system;

[0054] Z sets of second reference information are obtained by using the Z pieces of second error reporting information and the Z second problem solutions corresponding one-to-one to the Z pieces of second error reporting information.

[0055] In one example, after Z pieces of second error information are obtained, for each piece of the Z pieces of second error information, a second problem solution corresponding to the second error information may be retrieved from a knowledge base of the target system using a large language model.

[0056] After obtaining Z pieces of second error information and Z second problem solutions corresponding one to the Z pieces of second error information, the second error information and second problem solutions that have a corresponding relationship among the Z pieces of second error information and the Z second problem solutions can be combined into a group of second reference information, and finally, M groups of second reference information are obtained.

[0057] In the above example, Z pieces of second error information can be automatically selected from the multiple remaining error information (the multiple remaining error information are other candidate error information except the M first error information in the multiple candidate error information; the second error information is the error information that can retrieve the corresponding second problem solution in the knowledge base of the target system), and the Z pieces of second error information and the Z second problem solutions corresponding to the Z pieces of second error information are used to obtain Z groups of second reference information. In this way, the efficiency of obtaining the Z groups of second reference information can be improved to further improve the efficiency of solving system problems, thereby further improving the timeliness of operation and maintenance of the target system.

[0058] After obtaining M groups of first reference information and Z groups of second reference information, N groups of optimized reference information can be obtained based on the M groups of first reference information and the Z groups of second reference information. For example, the M groups of first reference information and the Z groups of second reference information can be used together as N groups of optimized reference information. At this time, N=M+Z. In another example, N groups of optimized reference information can also be obtained in the following manner:

[0059] Selecting multiple groups of preliminary reference information with the highest similarity to the current error message from the M groups of first reference information and the Z groups of second reference information;

[0060] Based on the multiple groups of preliminary reference information, N groups of optimized reference information are obtained.

[0061] In a specific example, for each group of first reference information in the M groups of first reference information, the similarity between the first error information carried in the first reference information and the current error information can be calculated. Exemplarily, the first error information carried in the first reference information can be vectorized to obtain a first vectorized processing result, and the current error information can be vectorized to obtain a second vectorized processing result, and the similarity between the first vectorized result and the second vectorized processing result can be calculated by a cosine similarity algorithm to characterize the similarity between the first error information carried in the first reference information and the current error information.

[0062] Similarly, for each group of second reference information in the Z groups of second reference information, the similarity between the second error information carried in the second reference information and the current error information can be calculated. Exemplarily, the second error information carried in the second reference information can be vectorized to obtain a third vectorized processing result, and the current error information can be vectorized to obtain a fourth vectorized processing result, and the similarity between the third vectorized result and the fourth vectorized processing result can be calculated by a cosine similarity algorithm to characterize the similarity between the second error information carried in the second reference information and the current error information.

[0063] Thereafter, multiple groups of reference information (including the first reference information and / or the second reference information) whose error information (the first error information carried in the first reference information or the second error information carried in the second reference information) has the highest similarity with the current error information can be selected from the M groups of first reference information and the Z groups of second reference information as multiple groups of preliminary reference information. The specific number of preliminary reference information can be set according to actual application requirements, and the embodiments of the present disclosure do not limit this.

[0064] After selecting multiple groups of preliminary reference information with the highest similarity to the current error information from the M groups of first reference information and the Z groups of second reference information, N groups of optimized reference information can be obtained in the following manner:

[0065] Obtaining a reliability score corresponding to each set of preliminary selected reference information in the plurality of sets of preliminary selected reference information;

[0066] N groups of preliminary selected reference information with the highest reliability scores are selected from multiple groups of preliminary selected reference information as optimized reference information to obtain N groups of optimized reference information.

[0067] Among them, for each group of preliminary reference information in the multiple groups of preliminary reference information, the reliability score corresponding to the preliminary reference information is the reliability score of the preliminary problem solution carried in the preliminary reference information, which is used to characterize the degree of match between the preliminary error information carried in the preliminary reference information and the preliminary problem solution. Here, when the preliminary problem solution carried in the preliminary reference information can solve the system problem described by the preliminary error information it carries, it characterizes that the preliminary error information carried in the preliminary reference information has a high degree of match with the preliminary problem solution, that is, it characterizes that the reliability score corresponding to the preliminary reference information is high; on the contrary, when the preliminary problem solution carried in the preliminary reference information cannot solve the system problem described by the preliminary error information it carries, it characterizes that the preliminary error information carried in the preliminary reference information has a low degree of match with the preliminary problem solution, that is, it characterizes that the reliability score corresponding to the preliminary reference information is low.

[0068] In a specific example, for each group of the multiple groups of preliminary selected reference information, the reliability score corresponding to the group of preliminary selected reference information (hereinafter referred to as the first target reference information for convenience of description) can be obtained in the following manner:

[0069] In a case where the first target reference information is derived from a system log of the target system, obtaining a reliability score corresponding to the first target reference information from the system log;

[0070] Alternatively, when the first target reference information comes from the knowledge base of the target system, the correlation score between the first target preliminary error information carried in the first target reference information and the first target problem solution is calculated to characterize the reliability score corresponding to the first target reference information.

[0071] In actual implementation, the Pearson correlation coefficient, Spearman rank correlation coefficient, Kendall rank correlation coefficient, etc. can be used to calculate the correlation score between the first target preliminary error information carried in the first target reference information and the first target problem solution, so as to characterize the reliability score corresponding to the first target reference information.

[0072] In another specific example, for each group of the multiple groups of preliminary selected reference information, the reliability score corresponding to the group of preliminary selected reference information (hereinafter referred to as the second target reference information for convenience of description) may be obtained in the following manner:

[0073] In the case where the second target reference information is derived from a system log of the target system, a reliability score corresponding to the second target reference information is obtained from the system log, and a relevance score between the second target error information carried in the second target reference information and the second target problem solution is calculated, and based on the reliability score corresponding to the second target reference information and the relevance score between the second target error information carried in the second target reference information and the second target problem solution, a reliability score corresponding to the second target reference information is obtained;

[0074] Alternatively, when the second target reference information comes from the knowledge base of the target system, the correlation score between the second target error information carried in the second target reference information and the second target problem solution is calculated to characterize the reliability score corresponding to the second target reference information.

[0075] In actual implementation, the Pearson correlation coefficient, Spearman rank correlation coefficient, Kendall rank correlation coefficient, etc. can be used to calculate the correlation score between the second target error information carried in the second target reference information and the second target problem solution.

[0076] In addition, in the above example, when the second target reference information originates from the system log of the target system, after obtaining the reliability score corresponding to the second target reference information from the system log and calculating the correlation score between the second target error information carried in the second target reference information and the second target problem solution, the reliability score corresponding to the second target reference information and the sum of the correlation scores between the second target error information carried in the second target reference information and the second target problem solution can be calculated as the reliability score corresponding to the second target reference information.

[0077] In the above example, multiple groups of preliminary reference information with the highest similarity to the current error information can be selected from M groups of first reference information and Z groups of second reference information, and N groups of optimized reference information can be obtained based on the multiple groups of preliminary reference information. In this way, when prompt information is obtained based on N groups of optimized reference information in the future, and the current problem solution for the current error information is obtained based on the prompt information using the large model, the data processing amount of the large model can be reduced to further improve the efficiency of solving system problems, thereby further improving the timeliness of operation and maintenance of the target system. In addition, when N groups of optimized reference information are obtained based on multiple groups of preliminary reference information, the reliability score corresponding to each group of preliminary reference information in the multiple groups of preliminary reference information is obtained, and the N groups of preliminary reference information with the highest reliability score are selected from the multiple groups of preliminary reference information as the optimized reference information to obtain N groups of optimized reference information. In this way, when prompt information is obtained based on N groups of optimized reference information in the future, and the current problem solution for the current error information is obtained based on the prompt information using the large model, the reliability of the output result of the large model (that is, the current problem solution) can also be improved.

[0078] Through the above method, in the embodiment of the present disclosure, based on the system log of the target system, M groups of first reference information (the first reference information includes the first error information related to the target system and the first problem solution for the first error information) can be obtained, and based on the knowledge base of the target system, Z groups of second reference information (the second reference information includes the second error information related to the target system and the second problem solution for the second error information) can be obtained, and based on the M groups of first reference information and the Z groups of second reference information, N groups of optimized reference information can be obtained, and then based on the N groups of optimized reference information, prompt information can be obtained. That is, through the above method, in the embodiment of the present disclosure, based on reference information from multiple sources (including M groups of first reference information from the system log of the target system and the second reference information from the knowledge base of the target system), N groups of optimized reference information can be obtained, and then based on the N groups of optimized reference information, prompt information can be obtained, thereby ensuring the richness and comprehensiveness of the prompt information, so as to further improve the reliability of the output result of the large model (that is, the current problem solution).

[0079] In some optional implementations, step S103, i.e., "using the large model and based on the prompt information, obtaining a current problem solution for the current error message" may include:

[0080] Utilize multiple large models and based on prompt information, obtain the current problem solution for the current error message.

[0081] Among them, the multiple large models may include ChatGPT, Claude 3.5Sonnet, Gemini 1.5Flash, Wenxinyiyan, etc. After using the multiple large models and obtaining the current problem solution for the current error message based on the prompt information, that is, obtaining the multiple current problem solutions corresponding to the multiple large models one by one, the multiple current problem solutions may be displayed respectively to provide users with more abundant choices, thereby enhancing the user experience.

[0082] In some optional implementations, after step S103, the method for obtaining a problem solution based on a large model further includes:

[0083] Get user evaluation results related to the current problem solution;

[0084] Based on the user evaluation results related to the current problem solution, a user recommendation score for the current problem solution is obtained.

[0085] In one example, a target problem solution related to the current problem solution and a user evaluation result related to the target problem solution may be obtained, and the user evaluation result related to the target problem solution may be used as the user evaluation result related to the current problem solution.

[0086] Among them, the target problem solution related to the current problem solution can be a problem solution for the target error information obtained based on the large model within the historical period. Here, the target error information has the same semantics as the current error information, that is, the target error information and the current error information point to the same system problem. In addition, in the above example, there can be multiple target problem solutions, and each of the multiple target problem solutions has a user evaluation result associated with it, that is, the user evaluation result related to the current problem solution includes multiple user evaluation results corresponding to multiple target problem solutions one by one.

[0087] Exemplarily, the user evaluation results related to the current problem solution include 5 user evaluation results corresponding to the 5 target problem solutions, namely user evaluation result A1, user evaluation result A2, user evaluation result A3, user evaluation result A4 and user evaluation result A5. Among them, user evaluation result A1 is "accurate answer", user evaluation result A2 is "accurate answer", user evaluation result A3 is "accurate answer", user evaluation result A4 is "inaccurate answer", and user evaluation result A5 is "accurate answer". Then, based on the user evaluation results related to the current problem solution, the user recommendation score of the current problem solution can be obtained: 4 / 5=0.8.

[0088] Furthermore, in the above implementation, after obtaining the user evaluation results related to the current problem solution and obtaining the user recommendation score of the current problem solution based on the user evaluation results, the problem solution acquisition method based on the large model further includes:

[0089] Based on the user recommendation score of the current problem solution, the reliability score of the current problem solution is obtained;

[0090] When the reliability score of the current problem solution does not meet the preset acceptance requirements, a manual processing request is generated based on the current error information and the current problem solution;

[0091] Send a manual processing request to the target device.

[0092] In one example, the reliability score of the current problem solution can be obtained by combining the user recommendation score of the current problem solution and the relevance score between the current error information and the current problem solution, so as to characterize the matching degree between the current error information and the current problem solution. When the current problem solution can solve the system problem described by the current error information, it characterizes that the current error information has a high matching degree with the current error information, that is, the reliability score characterizing the current problem solution is high; on the contrary, when the current problem solution cannot solve the system problem described by the current error information, it characterizes that the current error information has a low matching degree with the current error information, that is, the reliability score characterizing the current problem solution is low.

[0093] In a specific example, the Pearson correlation coefficient, Spearman rank correlation coefficient, Kendall rank correlation coefficient, etc. can be used to calculate the correlation score between the current error information and the current problem solution, and calculate the user recommendation score of the current problem solution, as well as the sum of the correlation scores of the current error information and the current problem solution as the reliability score of the current problem solution.

[0094] After obtaining the reliability score of the current problem solution, if the reliability score of the current problem solution is greater than or equal to the preset evaluation threshold, it can be determined that the reliability score of the current problem solution meets the preset acceptance requirement; if the reliability score of the current problem solution is less than the preset evaluation threshold, it can be determined that the reliability score of the current problem solution does not meet the preset acceptance requirement. The preset evaluation threshold can be set according to actual application requirements, and the embodiments of the present disclosure do not limit this.

[0095] If the reliability score of the current problem solution meets the preset acceptance requirements and the same problem solution exists in the system log of the target system, the reliability score of the same problem solution can be obtained from the system log, and if the reliability score of the same problem solution is lower than the reliability score of the current problem solution, the same problem solution in the system log of the target system is replaced by the current problem solution. The same problem solution and the current problem solution point to the same system problem.

[0096] In the case that the reliability score of the current problem solution does not meet the preset acceptance requirements, a manual processing request is generated based on the current error information and the current problem solution, and the manual processing request is sent to the target device to notify the target personnel to manually process the current error information. The target device can be a terminal device, such as a conventional computer, a tablet computer, a smart phone, a personal digital processing or other similar computing devices.

[0097] Through the above method, in the embodiment of the present disclosure, the user evaluation results related to the current problem solution can be obtained, and based on the user evaluation results related to the current problem solution, the user recommendation score of the current problem solution can be obtained for the user to choose, thereby enhancing the user experience. Moreover, based on the user recommendation score of the current problem solution, the reliability score of the current problem solution can be obtained, and when the reliability score of the current problem solution does not meet the preset acceptance requirements, a manual processing request is generated based on the current error information and the current problem solution, and then the manual processing request is sent to the target device to notify the target personnel to manually process the current error information, thereby further improving the timeliness of operation and maintenance of the target system.

[0098] The following will be combined Figure 2 and 3 , the complete process of a method for obtaining a problem solution based on a large model provided in an embodiment of the present disclosure is described.

[0099] Step S201, obtaining current error information of the target system.

[0100] Step S202, based on the system log of the target system, obtain M groups of first reference information; wherein M≥1 and is an integer; the first reference information includes first error information related to the target system and a first problem solution for the first error information.

[0101] Step S203, based on the knowledge base of the target system, obtain Z groups of second reference information; wherein Z ≥ 1 and is an integer; the second reference information includes second error information related to the target system and a second problem solution for the second error information.

[0102] Step S204: Obtain N groups of optimized reference information based on the M groups of first reference information and the Z groups of second reference information.

[0103] Step S205: Obtain prompt information based on N groups of optimization reference information.

[0104] Step S206, using the large model, based on the prompt information, obtain a current problem solution for the current error information.

[0105] Step S207: obtaining a user evaluation result related to the current problem solution, and obtaining a user recommendation score for the current problem solution based on the user evaluation result related to the current problem solution.

[0106] Step S208, based on the user recommendation score of the current problem solution, obtain the reliability score of the current problem solution, and if the reliability score of the current problem solution does not meet the preset acceptance requirements, generate a manual processing request based on the current error information and the current problem solution, and then send the manual processing request to the target device; wherein the reliability score of the current problem solution is used to characterize the degree of match between the current error information and the current problem solution.

[0107] For the specific functions and examples of each step in the embodiments of the present disclosure, please refer to the relevant descriptions of the corresponding steps in the aforementioned large model-based problem solution acquisition method embodiment, which will not be repeated here.

[0108] See also Figure 4 , which is a scenario diagram of a method for obtaining a problem solution based on a large model provided in an embodiment of the present disclosure.

[0109] As mentioned above, the method for obtaining a problem solution based on a large model provided by the embodiment of the present disclosure is applied to an electronic device, wherein the electronic device may be a server, a workbench, a mainframe computer, a conventional computer or other similar computing devices.

[0110] Electronic devices are used for:

[0111] Get the current error information of the target system;

[0112] Obtain prompt information; wherein the prompt information includes N groups of optimization reference information; N ≥ 1 and is an integer; the optimization reference information includes historical error information related to the target system and historical problem solutions for the historical error information;

[0113] Using the big model, based on the prompt information, we can get the current problem solution for the current error message.

[0114] It should be noted that in the embodiments of the present disclosure, Figure 4 The scene diagram shown is only illustrative and not restrictive. Those skilled in the art can Figure 4 Various obvious changes and / or substitutions are made to the examples, and the obtained technical solutions still fall within the scope of the disclosure of the embodiments of the present disclosure.

[0115] In order to better implement the problem solution acquisition method based on the large model, the embodiment of the present disclosure also provides a problem solution acquisition device based on the large model, which can be integrated in an electronic device. The electronic device can be a server, a workbench, a large computer, a conventional computer or other similar computing devices. Figure 5 The schematic structural block diagram shown illustrates a large model-based problem solution acquisition device 500 provided in the disclosed embodiment.

[0116] The problem solution acquisition device 500 based on the large model includes:

[0117] The error information acquisition unit 501 is used to acquire the current error information of the target system;

[0118] The prompt information acquisition unit 502 is used to acquire prompt information; wherein the prompt information includes N groups of optimization reference information; N ≥ 1 and is an integer; the optimization reference information includes historical error information related to the target system and historical problem solutions for the historical error information;

[0119] The problem solution obtaining unit 503 is used to obtain the current problem solution for the current error information based on the prompt information by using the large model.

[0120] In some optional implementations, the prompt information obtaining unit 502 is used to:

[0121] Based on the system log of the target system, M groups of first reference information are obtained; wherein M≥1 and is an integer; the first reference information includes first error information related to the target system and a first problem solution for the first error information;

[0122] Based on the knowledge base of the target system, Z groups of second reference information are obtained; wherein Z ≥ 1 and is an integer; the second reference information includes second error information related to the target system and a second problem solution for the second error information;

[0123] Based on the M groups of first reference information and the Z groups of second reference information, N groups of optimized reference information are obtained;

[0124] Based on N sets of optimization reference information, prompt information is obtained.

[0125] In some optional implementations, the prompt information obtaining unit 502 is used to:

[0126] Determine multiple candidate error messages from the system log of the target system;

[0127] Selecting M first error messages from a plurality of candidate error messages; wherein the first error messages are candidate error messages for which corresponding first problem solutions exist in the system log of the target system;

[0128] M groups of first reference information are obtained by using the M pieces of first error information and the M first problem solutions corresponding one-to-one to the M pieces of first error information.

[0129] In some optional implementations, the prompt information obtaining unit 502 is used to:

[0130] Selecting Z second error messages from the plurality of remaining error messages; wherein the plurality of remaining error messages are other candidate error messages except the M first error messages from the plurality of candidate error messages; the second error messages are error messages for which a corresponding second problem solution can be retrieved in a knowledge base of the target system;

[0131] Z sets of second reference information are obtained by using the Z pieces of second error reporting information and the Z second problem solutions corresponding one-to-one to the Z pieces of second error reporting information.

[0132] In some optional implementations, the prompt information obtaining unit 502 is used to:

[0133] Selecting multiple groups of preliminary reference information with the highest similarity to the current error message from the M groups of first reference information and the Z groups of second reference information;

[0134] Based on the multiple groups of preliminary reference information, N groups of optimized reference information are obtained.

[0135] In some optional implementations, the prompt information obtaining unit 502 is used to:

[0136] Obtaining a reliability score corresponding to each set of preliminary selection reference information in the plurality of sets of preliminary selection reference information; wherein, for each set of preliminary selection reference information in the plurality of sets of preliminary selection reference information, the reliability score corresponding to the preliminary selection reference information is used to characterize the degree of matching between the preliminary selection error information carried in the preliminary selection reference information and the preliminary selection problem solving device;

[0137] N groups of preliminary selected reference information with the highest reliability scores are selected from multiple groups of preliminary selected reference information as optimized reference information to obtain N groups of optimized reference information.

[0138] In some optional implementations, the large model-based problem solution acquisition device 500 further includes a user recommendation score acquisition unit, which is used to:

[0139] Get user evaluation results related to the current problem solution;

[0140] Based on the user evaluation results related to the current problem solution, a user recommendation score for the current problem solution is obtained.

[0141] In some optional implementations, the large model-based problem solution acquisition device 500 further includes a manual processing request unit, which is used to:

[0142] Based on the user recommendation score of the current problem solution, the reliability score of the current problem solution is obtained; wherein the reliability score of the current problem solution is used to characterize the matching degree between the current error information and the current problem solution;

[0143] When the reliability score of the current problem solution does not meet the preset acceptance requirements, a manual processing request is generated based on the current error information and the current problem solution;

[0144] Send a manual processing request to the target device.

[0145] In the embodiment of the present disclosure, the specific functions and examples of each unit in the large model-based problem solution acquisition device 500 can be found in the relevant descriptions of the corresponding steps in the aforementioned large model-based problem solution acquisition method embodiment, and will not be repeated here.

[0146] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0147] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0148] Figure 6A schematic block diagram of an example electronic device 600 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as vehicle-mounted computing devices, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0149] like Figure 6 As shown, the electronic device 600 includes a computing unit 601, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 602 or a computer program loaded from a storage unit 608 to a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0150] A number of components in the electronic device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of renderers, speakers, etc.; a storage unit 608, such as a disk, an optical disk, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the electronic device 600 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0151] The computing unit 601 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 601 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSP), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 601 performs the various methods and processes described above, for example, a method for obtaining a problem solution based on a large model. For example, in some embodiments, a method for obtaining a problem solution based on a large model may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, for example, a storage unit 608. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps in the method for obtaining a problem solution based on a large model described above may be executed. Alternatively, in other embodiments, the computing unit 601 may be configured as a large model-based problem solution acquisition method in any other appropriate manner (for example, by means of firmware).

[0152] Various embodiments of the systems and techniques described above herein may be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0153] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0154] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a RAM, a ROM, an erasable programmable read-only memory (EPROM) or a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0155] To provide interaction with a user, the systems and techniques described herein may be implemented on a computer having: a rendering device (e.g., a cathode ray tube (CRT) renderer or a liquid crystal display (LCD)) for rendering information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices are also used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0156] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: Local Area Network (LAN), Wide Area Network (WAN), and the Internet.

[0157] The computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server can be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0158] The embodiment of the present disclosure also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to enable a computer to execute a method for obtaining a problem solution based on a large model.

[0159] The embodiment of the present disclosure also provides a computer program product, including a computer program, which, when executed by a processor, implements a method for obtaining a problem solution based on a large model.

[0160] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in the present disclosure can be executed in parallel, sequentially or in a different order, as long as the desired results of the technical solutions disclosed in the present disclosure can be achieved, and this document is not limited here. In addition, in the present disclosure, relational terms such as "first", "second", "third", etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. In addition, "multiple" in the present disclosure can be understood as at least two.

[0161] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for obtaining a problem solution based on a large model, comprising: Get the current error information of the target system; Acquire prompt information; wherein the prompt information includes N groups of optimization reference information; N≥1 and is an integer; the optimization reference information includes historical error information related to the target system and historical problem solutions for the historical error information; By utilizing the large model and based on the prompt information, a current problem solution for the current error information is obtained.

2. The method according to claim 1, wherein: The obtaining of prompt information includes: Based on the system log of the target system, M groups of first reference information are obtained; wherein M≥1 and is an integer; the first reference information includes first error information related to the target system and a first problem solution for the first error information; Based on the knowledge base of the target system, Z groups of second reference information are obtained; wherein Z ≥ 1 and is an integer; the second reference information includes second error information related to the target system and a second problem solution for the second error information; Based on the M groups of first reference information and the Z groups of second reference information, obtaining the N groups of optimized reference information; The prompt information is obtained based on the N groups of optimization reference information.

3. The method according to claim 2, wherein: The obtaining, based on the system log of the target system, M groups of first reference information includes: Determining a plurality of candidate error messages from a system log of the target system; Selecting M first error messages from the plurality of candidate error messages; wherein the first error messages are candidate error messages for which corresponding first problem solutions exist in the system log of the target system; The M groups of first reference information are obtained by using the M pieces of first error information and the M first problem solutions corresponding one-to-one to the M pieces of first error information.

4. The method according to claim 3, wherein: The obtaining Z groups of second reference information based on the knowledge base of the target system includes: Selecting Z second error messages from the plurality of remaining error messages; wherein the plurality of remaining error messages are other candidate error messages among the plurality of candidate error messages except the M first error messages; the second error messages are error messages for which a corresponding second problem solution can be retrieved in the knowledge base of the target system; The Z groups of second reference information are obtained by using the Z pieces of second error information and the Z second problem solutions corresponding one-to-one to the Z pieces of second error information.

5. The method according to claim 2, wherein: The obtaining the N groups of optimized reference information based on the M groups of first reference information and the Z groups of second reference information includes: Selecting a plurality of groups of preliminary reference information having the highest similarity to the current error information from the M groups of first reference information and the Z groups of second reference information; Based on the multiple groups of preliminary reference information, the N groups of optimized reference information are obtained.

6. The method according to claim 5, wherein: The obtaining the N groups of optimized reference information based on the multiple groups of preliminary selected reference information includes: Obtaining a reliability score corresponding to each group of preliminary selected reference information in the multiple groups of preliminary selected reference information; wherein, for each group of preliminary selected reference information in the multiple groups of preliminary selected reference information, the reliability score corresponding to the preliminary selected reference information is used to characterize the degree of matching between the preliminary selected error information carried in the preliminary selected reference information and the preliminary selected problem solution; N groups of preliminary selected reference information with the highest reliability scores are selected from the multiple groups of preliminary selected reference information as optimized reference information to obtain the N groups of optimized reference information.

7. The method according to claim 1, further comprising: Obtaining user evaluation results related to the current problem solution; Based on the user evaluation results related to the current problem solution, a user recommendation score of the current problem solution is obtained.

8. The method according to claim 7, further comprising: Based on the user recommendation score of the current problem solution, a reliability score of the current problem solution is obtained; wherein the reliability score of the current problem solution is used to characterize the matching degree between the current error information and the current problem solution; If the reliability score of the current problem solution does not meet the preset acceptance requirement, a manual processing request is generated based on the current error information and the current problem solution; The manual processing request is sent to a target device.

9. A device for obtaining a problem solution based on a large model, comprising: An error information acquisition unit, used to acquire current error information of the target system; A prompt information acquisition unit, used to acquire prompt information; wherein the prompt information includes N groups of optimization reference information; N ≥ 1 and is an integer; the optimization reference information includes historical error information related to the target system and historical problem solutions for the historical error information; The problem solution obtaining unit is used to obtain a current problem solution for the current error information based on the prompt information by using the large model.

10. The device according to claim 9, wherein: The prompt information acquisition unit is used for: Based on the system log of the target system, M groups of first reference information are obtained; wherein M≥1 and is an integer; the first reference information includes first error information related to the target system and a first problem solution for the first error information; Based on the knowledge base of the target system, Z groups of second reference information are obtained; wherein Z ≥ 1 and is an integer; the second reference information includes second error information related to the target system and a second problem solution for the second error information; Based on the M groups of first reference information and the Z groups of second reference information, obtaining the N groups of optimized reference information; The prompt information is obtained based on the N groups of optimization reference information.

11. The device according to claim 10, wherein: The prompt information acquisition unit is used for: Determining a plurality of candidate error messages from a system log of the target system; Selecting M first error messages from the plurality of candidate error messages; wherein the first error messages are candidate error messages for which corresponding first problem solutions exist in the system log of the target system; The M groups of first reference information are obtained by using the M pieces of first error information and the M first problem solutions corresponding one-to-one to the M pieces of first error information.

12. The device according to claim 11, wherein The prompt information acquisition unit is used for: Selecting Z second error messages from the plurality of remaining error messages; wherein the plurality of remaining error messages are other candidate error messages among the plurality of candidate error messages except the M first error messages; the second error messages are error messages for which a corresponding second problem solution can be retrieved in the knowledge base of the target system; The Z groups of second reference information are obtained by using the Z pieces of second error information and the Z second problem solutions corresponding one-to-one to the Z pieces of second error information.

13. The device according to claim 10, wherein: The prompt information acquisition unit is used for: Selecting a plurality of groups of preliminary reference information having the highest similarity to the current error information from the M groups of first reference information and the Z groups of second reference information; Based on the multiple groups of preliminary reference information, the N groups of optimized reference information are obtained.

14. The device according to claim 13, wherein: The prompt information acquisition unit is used for: Obtaining a reliability score corresponding to each group of preliminary selection reference information in the multiple groups of preliminary selection reference information; wherein, for each group of preliminary selection reference information in the multiple groups of preliminary selection reference information, the reliability score corresponding to the preliminary selection reference information is used to characterize the degree of matching between the preliminary selection error information carried in the preliminary selection reference information and the preliminary selection problem solving device; N groups of preliminary selected reference information with the highest reliability scores are selected from the multiple groups of preliminary selected reference information as optimized reference information to obtain the N groups of optimized reference information.

15. The device according to claim 9, further comprising a user recommendation score acquisition unit, configured to: Obtaining user evaluation results related to the current problem solution; Based on the user evaluation results related to the current problem solution, a user recommendation score of the current problem solution is obtained.

16. The apparatus according to claim 15, further comprising a manual processing request unit, configured to: Based on the user recommendation score of the current problem solution, the reliability score of the current problem solution is obtained; wherein, The reliability score of the current problem solution is used to represent the matching degree between the current error information and the current problem solution; If the reliability score of the current problem solution does not meet the preset acceptance requirement, a manual processing request is generated based on the current error information and the current problem solution; The manual processing request is sent to a target device.

17. An electronic device comprising: at least one processor; a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method according to any one of claims 1 to 8.

18. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1 to 8.

19. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the method according to any one of claims 1 to 8.

Images