Abnormal data reproduction method and abnormal code reproduction method

Abstract

The embodiments of this specification provide an abnormal data reproduction method and an abnormal code reproduction method, wherein the abnormal data reproduction method comprises: obtaining abnormal description information of target abnormal data; extracting target reproduction experience corresponding to the target abnormal data from a reproduction experience library, wherein the reproduction experience library comprises multiple reproduction experiences, and the reproduction experience is used to guide the reproduction process of the abnormal data; and reproducing the target abnormal data according to the abnormal description information and the target reproduction experience to obtain the reproduction data. By introducing the target reproduction experience to automatically reproduce the target abnormal data, the efficiency and success rate of abnormal data reproduction are improved.

Description

Technical Field

[0001] The embodiments of the present specification relate to the field of computer technology, and in particular to a method for reproducing abnormal data and a method for reproducing abnormal code. Background Art

[0002] With the development of computer technology, automated data processing has become more and more mature. In the process of data processing, there are usually some abnormal data. If these abnormal data can be successfully reproduced, it can help identify the root cause of the problem, help find system bottlenecks or performance problems, and thus perform targeted optimization. Taking the reproduction of problem reports or feature requests (Issues) in software development as an example, the successful reproduction of Issues can help locate faults and troubleshoot (debug) and repair programs.

[0003] Currently, abnormal data is usually reproduced manually by staff. However, manual reproduction of abnormal data requires a lot of manpower and introduces uncertain factors, resulting in a high failure rate of abnormal data reproduction and extremely low reproduction efficiency. Therefore, an efficient and accurate abnormal data reproduction solution is urgently needed. Summary of the invention

[0004] In view of this, an embodiment of this specification provides an abnormal data reproduction method. One or more embodiments of this specification also relate to an abnormal code reproduction method, an abnormal data reproduction device, an abnormal code reproduction device, a computing device, a computer-readable storage medium and a computer program product to solve the technical defects existing in the prior art.

[0005] According to a first aspect of an embodiment of this specification, a method for reproducing abnormal data is provided, comprising:

[0006] Get the abnormal description information of the target abnormal data;

[0007] Extracting target recurrence experience corresponding to target abnormal data from the recurrence experience database, wherein the recurrence experience database includes multiple recurrence experiences, and the recurrence experience is used to guide the recurrence process of abnormal data;

[0008] Reproduce the target abnormal data based on the abnormal description information and target reproduction experience to obtain the reproduction data.

[0009] According to a second aspect of an embodiment of this specification, a method for reproducing an abnormal code is provided, comprising:

[0010] Get the exception description information of the target exception code;

[0011] Extracting target recurrence experience corresponding to target abnormal code from the recurrence experience library, wherein the recurrence experience library includes multiple recurrence experiences, and the recurrence experience is used to guide the recurrence process of the abnormal code;

[0012] Reproduce the target exception code based on the exception description information and target reproduction experience to obtain the reproduction code.

[0013] According to a third aspect of an embodiment of this specification, there is provided an abnormal data reproduction device, including:

[0014] A first acquisition module is configured to acquire abnormal description information of target abnormal data;

[0015] A first extraction module is configured to extract target recurrence experience corresponding to target abnormal data from a recurrence experience library, wherein the recurrence experience library includes a plurality of recurrence experiences, and the recurrence experience is used to guide the recurrence process of the abnormal data;

[0016] The first reproduction module is configured to reproduce target abnormal data according to the abnormal description information and target reproduction experience to obtain reproduction data.

[0017] According to a fourth aspect of an embodiment of this specification, there is provided an abnormal code reproduction device, including:

[0018] A second acquisition module is configured to acquire exception description information of a target exception code;

[0019] A second extraction module is configured to extract a target recurrence experience corresponding to a target abnormal code from a recurrence experience library, wherein the recurrence experience library includes a plurality of recurrence experiences, and the recurrence experience is used to guide the recurrence process of the abnormal code;

[0020] The second reproduction module is configured to reproduce the target exception code according to the exception description information and the target reproduction experience to obtain the reproduction code.

[0021] According to a fifth aspect of an embodiment of this specification, a computing device is provided, including:

[0022] Memory and processor;

[0023] The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor, the steps of the method provided in the first aspect or the second aspect are implemented.

[0024] According to the sixth aspect of the embodiments of this specification, a computer-readable storage medium is provided, which stores a computer program / instruction, and when the computer program / instruction is executed by a processor, the steps of the method provided in the first aspect or the second aspect are implemented.

[0025] According to a seventh aspect of the embodiments of this specification, a computer program product is provided, including a computer program / instruction, which, when executed by a processor, implements the steps of the method provided in the first aspect or the second aspect above.

[0026] An abnormal data reproduction method provided by an embodiment of the present specification includes: obtaining abnormal description information of target abnormal data; extracting target reproduction experience corresponding to the target abnormal data from a reproduction experience library, wherein the reproduction experience library includes multiple reproduction experiences, and the reproduction experience is used to guide the reproduction process of the abnormal data; according to the abnormal description information and the target reproduction experience, the target abnormal data is reproduced to obtain the reproduction data. By introducing the target reproduction experience to automatically reproduce the target abnormal data, the efficiency and success rate of abnormal data reproduction are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 This is an architecture diagram of an abnormal data reproduction system provided by an embodiment of this specification;

[0028] Figure 2 is a flow chart of an abnormal data reproduction method provided by an embodiment of this specification;

[0029] Figure 3 This is a flow chart of an abnormal code reproduction method provided by an embodiment of this specification;

[0030] Figure 4 It is a process flow chart of a method for reproducing abnormal codes provided by an embodiment of this specification;

[0031] Figure 5 It is a structural schematic diagram of an abnormal data reproduction device provided by an embodiment of this specification;

[0032] Figure 6 It is a structural schematic diagram of an abnormal code reproduction device provided by an embodiment of this specification;

[0033] Figure 7 It is a structural block diagram of a computing device provided by an embodiment of this specification. DETAILED DESCRIPTION

[0034] Many specific details are described in the following description to facilitate a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of this specification, so this specification is not limited to the specific implementation disclosed below.

[0035] The terms used in one or more embodiments of this specification are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of this specification. The singular forms of "a", "said" and "the" used in one or more embodiments of this specification and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0036] It should be understood that although the terms first, second, etc. may be used to describe various information in one or more embodiments of this specification, this information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of this specification, the first may also be referred to as the second, and similarly, the second may also be referred to as the first. Depending on the context, the word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining".

[0037] In addition, it should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in one or more embodiments of this specification are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0038] In one or more embodiments of this specification, a large model refers to a deep learning model with large-scale model parameters, which usually contains hundreds of millions, tens of billions, hundreds of billions, trillions, or even more than 10 trillion model parameters. A large model can also be called a foundation model / foundation model. The large model is pre-trained with large-scale unlabeled corpus to produce a pre-trained model with more than 100 million parameters. This model can adapt to a wide range of downstream tasks, and the model has good generalization ability, such as a large-scale language model (LLM, Large Language Model), a multi-modal pre-training model, etc.

[0039] When the big model is used in practice, only a small number of samples are needed to fine-tune the pre-trained model and it can be applied to different tasks. The big model can be widely used in natural language processing (NLP), computer vision and other fields. Specifically, it can be applied to computer vision tasks such as visual question answering (VQA), image description (IC, Image Caption), image generation, as well as natural language processing tasks such as text-based sentiment classification, text summary generation, and machine translation. The main application scenarios of the big model include digital assistants, intelligent robots, search, online education, office software, e-commerce, intelligent design, etc.

[0040] First, the terms involved in one or more embodiments of this specification are explained.

[0041] Problem report: refers to a file that contains enough information to help developers understand and reproduce the problem. The problem report includes the specific steps in which the error occurred, the environment used (operating system version, browser, etc.), and related log files or screenshots. Problem reports can help developers quickly locate and fix problems, thereby improving the quality and stability of the software.

[0042] Feature Request: A request from a user or stakeholder to add a new feature to the software or to improve an existing feature.

[0043] Reproduction: refers to the process of reproducing the problem in the same environment.

[0044] Software Engineering Benchmark (SWE-bench): refers to a benchmark test set that evaluates the performance of automated software engineering systems and is used to evaluate the benchmark for models to solve real-world software problems.

[0045] Taking the reproduction of abnormal data in software development scenarios as an example, an issue usually contains the core code or steps to reproduce the problem. To solve the issue, the programmer needs to reproduce the error, modify the code, and then run the reproduced code again for verification. Therefore, issue reproduction is an important step in achieving automatic software engineering.

[0046] The embodiment of this specification proposes an abnormal data reproduction solution based on reproduction experience, aiming to solve the problem that abnormal data (such as Issue) is difficult to reproduce. Specifically, the abnormal description information of the target abnormal data is obtained; the target reproduction experience corresponding to the target abnormal data is extracted from the reproduction experience library, wherein the reproduction experience library includes multiple reproduction experiences, and the reproduction experience is used to guide the reproduction process of the abnormal data; according to the abnormal description information and the target reproduction experience, the target abnormal data is reproduced to obtain the reproduction data.

[0047] In this specification, an abnormal data reproduction method is provided. This specification also relates to an abnormal code reproduction method, an abnormal data reproduction device, an abnormal code reproduction device, a computing device, a computer-readable storage medium and a computer program product, which are described in detail one by one in the following embodiments.

[0048] See also Figure 1 , Figure 1 An architecture diagram of an abnormal data reproduction system provided by an embodiment of the present specification is shown. The abnormal data reproduction system may include a client 100 and a server 200;

[0049] The client 100 is used to send abnormal description information of target abnormal data to the server 200;

[0050] The server 200 is used to extract the target reproduction experience corresponding to the target abnormal data from the reproduction experience library, wherein the reproduction experience library includes multiple reproduction experiences, and the reproduction experience is used to guide the reproduction process of the abnormal data; according to the abnormal description information and the target reproduction experience, the target abnormal data is reproduced to obtain the reproduction data.

[0051] In an optional embodiment of the present specification, the server 200 is further used to send the recurrence data to the client 100; the client 100 is further used to receive the recurrence data sent by the server 200.

[0052] By applying the solution of the embodiment of this specification, target abnormal data is automatically reproduced by introducing target reproduction experience, thereby improving the efficiency and success rate of abnormal data reproduction.

[0053] In actual applications, the abnormal data reproduction system may include multiple clients 100 and a server 200, wherein the client 100 may include an end-side device, and the server 200 may include a cloud-side device. A communication connection may be established between multiple clients 100 through the server 200. In the abnormal data reproduction scenario, the server 200 is used to provide abnormal data reproduction services between multiple clients 100. Multiple clients 100 may serve as senders or receivers, respectively, and achieve communication through the server 200. Users may interact with the server 200 through the client 100 to receive data sent by other clients 100, or send data to other clients 100, etc. In the abnormal data reproduction scenario, the user may publish a data stream to the server 200 through the client 100, and the server 200 may generate target reproduction data based on the data stream, and push the target reproduction data to other clients that have established communication. See. Figure 1 , the client 100 can also interact with the user through a graphical user interface to call the abnormal data reproduction service, thereby implementing the abnormal data reproduction method provided in the embodiment of this specification. Among them, the client 100 and the server 200 are connected through a network. The network provides a medium for the communication link between the client 100 and the server 200. The network can include various connection types, such as wired, wireless communication links or optical fiber cables, etc. The data transmitted by the client 100 may need to be encoded, transcoded, compressed, etc. before being released to the server 200.

[0054] The client 100 can be a browser, an APP (Application), or a web application such as an H5 (HyperText Markup Language 5, Hypertext Markup Language Version 5) application, or a light application (also known as a mini-program, a lightweight application) or a cloud application, etc. The client 100 can be based on the software development kit (SDK, Software Development Kit) of the corresponding service provided by the server 200, such as based on the real-time communication (RTC, Real Time Communication) SDK development and acquisition. The client 100 can be deployed in an electronic device and needs to rely on the device to run or some APPs in the device to run. For example, the electronic device can have a display screen and support information browsing, such as a personal mobile terminal such as a mobile phone, a tablet computer, a personal computer, etc. Various other types of applications can also be configured in the electronic device, such as human-computer dialogue applications, model training applications, text processing applications, web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0055] The server 200 may include servers that provide various services, such as servers that provide communication services to multiple clients, servers for background training that support models used on clients, and servers that process data sent by clients. It should be noted that the server 200 can be implemented as a distributed server cluster consisting of multiple servers, or as a single server. The server can also be a server of a distributed system, or a server combined with a blockchain. The server can also be a cloud server for basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content distribution networks (CDN, Content Delivery Network), and big data and artificial intelligence platforms, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology.

[0056] It is worth noting that the abnormal data reproduction method provided in the embodiments of this specification is generally performed by the server, but in other embodiments of this specification, the client may also have similar functions to the server, thereby performing the abnormal data reproduction method provided in the embodiments of this specification. In other embodiments, the abnormal data reproduction method provided in the embodiments of this specification can also be performed jointly by the client and the server. Next, taking the server executing the abnormal data reproduction method provided in the embodiments of this specification as an example, the abnormal data reproduction method is described in detail.

[0057] See also Figure 2 , Figure 2 A flowchart of an abnormal data reproduction method provided by an embodiment of the present specification is shown, which specifically includes the following steps:

[0058] Step 202: Obtain abnormal description information of target abnormal data.

[0059] It should be noted that target abnormal data refers to data with abnormal content, behavior or status. Target abnormal data can be abnormal problems or error instances that occur in the software system. These data may be caused by code errors, improper configuration, inconsistent data, changes in the external environment, etc., such as functional abnormal data, performance abnormal data, security vulnerability abnormalities, etc. Target abnormal data can also be erroneous data or inconsistent data in the text, such as spelling error data, grammatical error data, logical error data, format error data, etc. Abnormal description information refers to information used to describe the target abnormal data in detail. This information usually includes but is not limited to environmental information, scope of impact, error location, error type, and is selected according to actual conditions. The embodiments of this specification do not make any restrictions on this.

[0060] Exemplarily, assume that the target abnormal data is an abnormal problem in the text, and the abnormal description information of the target abnormal data can be "In the first sentence of the second paragraph on the first page, there is a problem that '发' is misspelled as '法'". Assume that the target abnormal data is an abnormal problem in the code, and the abnormal description information can be "On the user management platform, when a user attempts to change the password, the system always returns an error message [Password change failed, please try again]. The user reports that they have tried multiple times without success".

[0061] In practical applications, there are various ways to obtain the abnormal description information of the target abnormal data, which are specifically selected according to the actual situation, and the embodiments of this specification do not make any limitations on this. In one possible implementation manner of this specification, the abnormal description information of the target abnormal data sent by the user through the client can be received. In another possible implementation manner of this specification, the abnormal description information of the target abnormal data can be read from other data acquisition devices or databases.

[0062] Step 204: Extract the target reproduction experience corresponding to the target abnormal data from the reproduction experience library, where the reproduction experience library includes multiple reproduction experiences, and the reproduction experience is used to guide the reproduction process of the abnormal data.

[0063] It should be noted that the reproduction experience library is a collection used to store and manage the experiences and knowledge accumulated during the reproduction of abnormal problems (such as code Issues, system failures, data anomalies, etc.). The reproduction experience library usually contains a series of detailed steps on how to reproduce the abnormal problem, relevant environment configurations, known solutions, and other information helpful for diagnosing and solving the abnormal problem. There can be one or multiple reproduction experience libraries. For example, the reproduction experience library can be an experience library with a two-layer structure including a general experience library and a specific experience library. Among them, the general experience library is used to store reproduction techniques and patterns with strong universality, such as the experience library jointly maintained by multiple code repositories. The specific experience library is used to store unique reproduction experiences for specified tasks, such as the experience library maintained by a certain code repository. By setting up a reproduction experience library based on a two-layer structure, the universality and specificity of the experience can be taken into account at the same time, making the guiding ability of the target reproduction experience stronger. The target reproduction experience refers to at least one reproduction experience in the reproduction experience library applicable to the reproduction process of the target abnormal data. The reproduction experience library includes multiple reproduction experiences, and these multiple reproduction experiences can guide the reproduction process of at least one abnormal data. Different reproduction experiences can guide the reproduction process of the same abnormal data or different abnormal data, which are specifically selected according to the actual situation, and the embodiments of this specification do not make any limitations on this.

[0064] In practical applications, there are many ways to extract the target recurrence experience corresponding to the target abnormal data from the recurrence experience library, which can be selected according to the actual situation, and the embodiments of this specification do not limit this. In one possible implementation of this specification, all recurrence experiences in the recurrence experience library can be determined as the target recurrence experience corresponding to the target abnormal data.

[0065] In another possible implementation of the present specification, since the reproduction experience library may include a large number of reproduction experiences, determining all reproduction experiences as target reproduction experiences may cause the subsequent reproduction process of target abnormal data to consume a lot of time or even be unable to be completed. Therefore, the target reproduction experience can be screened out from the reproduction experience library according to the abnormal description information. That is, the above-mentioned extraction of the target reproduction experience corresponding to the target abnormal data from the reproduction experience library may include the following steps:

[0066] Obtain description information corresponding to multiple recurrence experiences in the recurrence experience library;

[0067] Matching the abnormal description information with the description information to obtain a matching result, wherein the matching result corresponds to the description information one by one;

[0068] According to the matching results, the target reproduction experience corresponding to the target abnormal data is extracted from the multiple reproduction experiences.

[0069] It should be noted that the matching result is used to describe the degree of match between the abnormal description information and the description information. The matching result can be a specific matching value, such as a matching degree of 80%, or a specific matching level, such as a very good match, a relatively good match, or a mismatch. There are many ways to obtain the description information corresponding to multiple recurrence experiences in the recurrence experience library, which are selected according to the actual situation. The embodiments of this specification do not impose any restrictions on this. In one possible implementation of this specification, the recurrence experience library can store not only the recurrence experience, but also the description information corresponding to each recurrence experience, which is used to describe the abnormal problems of the abnormal data corresponding to the recurrence experience. Therefore, the description information corresponding to multiple recurrence experiences can be read directly from the recurrence experience library. In another possible implementation of this specification, the description information corresponding to each recurrence experience can be matched from other data acquisition devices or databases.

[0070] In actual applications, the exception description information and the description information are matched. When the matching result is obtained, the similarity between the exception description information and the description information can be calculated, and the matching result can be determined based on the similarity calculation result. Among them, the similarity calculation method includes but is not limited to the cosine similarity calculation method and the Euclidean distance calculation method. The specific selection is based on the actual situation, and the embodiments of this specification do not impose any limitations on this.

[0071] Further, after obtaining the matching result between the abnormal description information and the description information, the recurrence experience corresponding to the description information whose matching result satisfies the matching condition can be determined as the target recurrence data corresponding to the target abnormal data. The matching condition can be that the matching degree is greater than a preset degree, such as a matching value greater than a preset threshold of 90%, or that the ranking is within the top N, where N is a positive integer, and the matching results with a higher matching degree are ranked before the matching results with a lower matching degree.

[0072] By applying the solution of the embodiment of this specification, the target reproduction experience is extracted from multiple reproduction experiences based on the matching results between the exception description information and the description information, thereby ensuring the adaptability between the target reproduction experience and the exception description information of the target exception data, and further improving the exception reproduction efficiency.

[0073] In an optional embodiment of the present specification, before extracting the target reproduction experience corresponding to the target abnormal data from the reproduction experience library, the reproduction experience library can also be constructed based on the abnormal data description information and the reproduction data of the abnormal data. That is, before extracting the target reproduction experience corresponding to the target abnormal data from the reproduction experience library, the following steps can also be included:

[0074] Obtain description information of abnormal data and reproduction data of abnormal data;

[0075] Input the description information and the reproduction data into the data processing model to obtain the reproduction results and reproduction experience of the reproduction data, wherein the reproduction results are used to describe the reproduction effect of the reproduction data;

[0076] Build a reproduction experience library based on the reproduction results and reproduction experience.

[0077] It should be noted that the reproduction data refers to the data used to reproduce the abnormal problems existing in the abnormal data. The abnormal problems existing in the abnormal data can be reproduced through the reproduction data, so that researchers or engineers can further analyze or solve the problems. The data processing model refers to a model with the ability to judge the reproduction effect and extract experience. The data processing model can be a pre-trained large model or a domain large model, or it can be a deep learning model trained based on sample description information, sample reproduction data, sample reproduction results, and sample reproduction experience. The reproduction results include successful reproduction and failed reproduction. If the reproduction result is a successful reproduction, it means that the abnormal problems existing in the abnormal data can be successfully reproduced through the reproduction data, and the reproduction experience obtained from the description information and the reproduction data can be regarded as a positive example in the abnormal reproduction process; if the reproduction result is a failed reproduction, it means that the abnormal problems existing in the abnormal data cannot be successfully reproduced through the reproduction data, and the reproduction experience obtained from the description information and the reproduction data can be regarded as a negative example in the abnormal reproduction process.

[0078] In practical applications, there are many ways to obtain the description information of abnormal data and the reproduction data of abnormal data, which can be selected according to the actual situation, and the embodiments of this specification do not impose any restrictions on this. In one possible implementation of this specification, the description information of abnormal data and the reproduction data of abnormal data sent by the user through the client can be received. In another possible implementation of this specification, the description information of abnormal data and the reproduction data of abnormal data can be extracted from other data acquisition devices or databases.

[0079] Furthermore, when constructing a reproduction experience library based on the reproduction results and the reproduction experience, in one possible implementation of this specification, the reproduction experience can be divided into positive examples and negative examples based on the reproduction results, so that the reproduction experience can be stored in the reproduction experience library by category, wherein the reproduction experience library can be an empty database or a database storing other reproduction experiences or other data before storing the reproduction experience. In another possible implementation of this specification, when storing the reproduction experience by category based on the reproduction results, the description information of the abnormal data and the reproduction data can also be additionally stored to obtain the reproduction experience library.

[0080] In an optional embodiment of the present specification, the reproduction experience library may include a first reproduction experience library and a second reproduction experience library, wherein the first reproduction experience library is used to store general reproduction experience of different tasks. The second reproduction experience library is used to store exclusive reproduction experience of a specified task. When extracting the target reproduction experience corresponding to the target abnormal data from the reproduction experience library, the general target reproduction experience can be extracted from the first reproduction experience library, or the target reproduction experience can be extracted from the second reproduction experience library used to store the exclusive reproduction experience of the task to which the target abnormal data belongs. The target reproduction experience can also be extracted from the first reproduction experience library and the second reproduction experience library at the same time. Further, according to the reproduction result and the reproduction experience, when constructing the reproduction experience library, it can be determined whether the abnormal data belongs to the same task as the target abnormal data. If so, the reproduction experience is stored in the second reproduction experience library based on the reproduction result. If not, the reproduction experience is stored in the first reproduction experience library based on the reproduction result.

[0081] By applying the solution of the embodiments of this specification, the reproduction experience and reproduction results are obtained based on the description information of the abnormal data and the corresponding reproduction data through the data processing model, and a reproduction experience library is constructed according to the reproduction results and reproduction experience, thereby obtaining a reproduction experience library that can provide abnormal data reproduction experience, laying the foundation for subsequent abnormal data reproduction.

[0082] In actual applications, when the description information and reproduction data are input into the data processing model to obtain the reproduction results and reproduction experience of the reproduction data, in a possible implementation method of this specification, the description information and reproduction data of the abnormal data can be directly input into the data processing model to obtain the reproduction results and reproduction experience of the reproduction data.

[0083] In another possible implementation of the present specification, the original recurrence experience in the original experience library may be additionally input into the data processing model, so that the data processing model outputs the recurrence result and the recurrence experience and additionally outputs the experience update suggestion of the recurrence experience. That is, the above-mentioned inputting the description information and the recurrence data into the data processing model to obtain the recurrence result and the recurrence experience of the recurrence data may include the following steps:

[0084] Inputting the original recurrence experience, description information and recurrence data in the original experience library into the data processing model, obtaining the recurrence result of the recurrence data, the recurrence experience and the experience update suggestion, wherein the experience update suggestion includes at least one of adding suggestion, deleting suggestion, modifying suggestion and merging suggestion;

[0085] Based on the reproduction results and reproduction experience, building a reproduction experience library can include the following steps:

[0086] According to the experience update suggestions, reproduction results and reproduction experience, the original experience library is updated to obtain the reproduction experience library.

[0087] It should be noted that the original experience library is used to store multiple original reproduction experiences. The original experience library can be one or more. Experience update suggestions refer to suggestions for updating reproduction experiences in the original experience library. Add suggestions indicate that the reproduction experience is an experience that is not in the original experience library, and it is recommended that the server add new reproduction experiences to the original experience library. Deletion suggestions indicate that the reproduction experience is an experience that already exists in the original experience library or is a new version or correct version of an experience that already exists in the original experience library. It is recommended that the server delete the outdated or erroneous original reproduction experience in the original experience library and store the reproduction experience in the original experience library. Modification suggestions indicate suggestions for modifying the original reproduction experience corresponding to the reproduction experience, such as editing, improving the original reproduction experience, agreeing with the original reproduction experience, and increasing the weight of the original reproduction experience, etc. Merge suggestions indicate that there are original reproduction experiences similar to the reproduction experience in the original experience library, and it is recommended to integrate similar original reproduction experiences with the reproduction experience, thereby improving the refinement of knowledge in the original experience library.

[0088] In actual applications, when the original experience library is updated based on experience update suggestions, reproduction results, and reproduction experience, the server can determine the experience update operation corresponding to the experience update suggestion, where the experience update operation includes at least one of addition, deletion, modification, and merging; based on the experience update operation and reproduction results, the reproduced experience is updated to the original experience library to obtain the reproduced experience library.

[0089] In an optional embodiment of the present specification, after inputting the description information and the recurrence data into the data processing model and obtaining the recurrence result and recurrence experience of the recurrence data, the original recurrence experience and the recurrence experience in the original experience library can be directly compared to obtain experience update suggestions. Of course, it is also possible to generate experience update suggestions for the original recurrence experience based only on the storage information of the original recurrence experience in the original experience library. For example, it is possible to determine whether to remove the original recurrence experience based on the last use time and / or use frequency of the original recurrence experience.

[0090] By applying the solution of the embodiment of this specification, the recurring experience is updated to the original experience library according to the experience update suggestion, thereby avoiding the storage of redundant experience in the original experience library, thereby improving the refinement and effectiveness of the data in the original experience library.

[0091] Step 206: Reproduce target abnormal data according to the abnormal description information and target reproduction experience to obtain reproduction data.

[0092] It should be noted that reproduction refers to the process of reproducing the target abnormal data corresponding to the abnormal description information through a series of predefined steps or conditions, so that the abnormal problem in the target abnormal data reappears. The purpose of reproduction is to verify the existence of the abnormal problem in the target abnormal data and provide a basis for further diagnosis and resolution of the abnormal problem. Reproduction data refers to the reproduction data obtained by reproducing the target abnormal data for the first time based on the abnormal description information and the target reproduction experience.

[0093] In practical applications, there are many ways to reproduce the target abnormal data and obtain the reproduced data according to the abnormal description information and the target reproduction experience, which are selected according to the actual situation. The embodiments of this specification do not impose any restrictions on this. In one possible implementation of this specification, the target abnormal data can be directly reproduced according to the abnormal description information and the target reproduction experience to obtain the reproduced data. In another possible implementation of this specification, the context data of the target abnormal data can also be additionally introduced to improve the accuracy of the reproduction process.

[0094] In an optional embodiment of the present specification, the above method of reproducing the target abnormal data according to the abnormal description information and the target reproduction experience, before obtaining the reproduction data, may further include the following steps:

[0095] Obtaining context data of target abnormal data;

[0096] Reproducing the target abnormal data based on the abnormal description information and target reproduction experience, and obtaining the reproduction data may include the following steps:

[0097] Reproduce the target abnormal data based on the context data, abnormal description information, and target reproduction experience to obtain the reproduction data.

[0098] It should be noted that the context data of the target exception data refers to the relevant background information and environmental data when the target exception data occurs. The context data can provide background information when the target exception data occurs, helping the server to better understand the overall picture of the exception problem, so as to more effectively diagnose and solve the exception problem. There are many ways to obtain the context data of the target exception data, which are selected according to the actual situation. The embodiments of this specification do not impose any restrictions on this. In one possible implementation of this specification, the context data of the target exception data sent by the user through the client can be received. In another possible implementation of this specification, the context data of the target exception data can be read from other data acquisition devices or databases.

[0099] In actual applications, the target abnormal data is reproduced according to the abnormal description information and the target reproduction experience. When the reproduction data is obtained, the reproduction prompt information, the abnormal description information and the target reproduction experience can be input into the data processing model, and the data processing model reproduces the target abnormal data to obtain the reproduction data. Among them, the reproduction prompt information is used to guide the data processing model to reproduce the abnormal data, such as "You are a professional programmer, please reproduce the target abnormal data according to the abnormal description information and the target reproduction experience, and output the reproduction data." According to the context data, the abnormal description information and the target reproduction experience, the target abnormal data is reproduced. When the reproduction data is obtained, the reproduction prompt information, the context data, the abnormal description information and the target reproduction experience can be input into the data processing model to obtain the reproduction data.

[0100] By applying the solution of the embodiments of this specification, in the process of reproducing target abnormal data according to abnormal description information and target reproduction experience, additional context data is introduced, thereby enhancing the understanding of the abnormal description information and improving the reproduction quality of the reproduced data.

[0101] In an optional embodiment of the present specification, the target abnormal data is reproduced according to the abnormal description information and the target reproduction experience. After the reproduction data is obtained, it can be determined whether the reproduction data can successfully reproduce the abnormal problem in the target abnormal data. In the case where the reproduction data can successfully reproduce the abnormal problem in the target abnormal data, the reproduction data is directly determined as the final target reproduction data. In the case where the reproduction data cannot successfully reproduce the abnormal problem in the target abnormal data, the reproduction data can be adjusted or the abnormal data can be reproduced again to obtain the target reproduction data that can successfully reproduce the abnormal problem in the target abnormal data. That is, the above-mentioned reproduction of the target abnormal data according to the abnormal description information and the target reproduction experience, after the reproduction data is obtained, can also include the following steps:

[0102] When the recurrence description information of the recurrence data matches the abnormal description information, the recurrence data is determined as the target recurrence data.

[0103] It should be noted that after obtaining the reproduction data, in order to prevent the reproduction data from directly outputting abnormal information instead of truly reproducing the abnormal problem, it is possible to determine whether the reproduction data has successfully reproduced the abnormal problem in the target abnormal data. The method for determining whether the initial reproduction has successfully reproduced the target abnormal data can be to determine whether the reproduction description information of the reproduction data matches the abnormal description information. If they match, it indicates that the reproduction data can successfully reproduce the abnormal problem in the target abnormal data, and the reproduction data can be directly determined as the target reproduction data. If they do not match, it means that the reproduction data cannot successfully reproduce the abnormal problem in the target abnormal data.

[0104] In practical applications, there are various ways to determine whether the reproduction description information matches the abnormal description information, which are specifically selected according to the actual situation, and the embodiments of this specification do not make any limitations in this regard. In one possible implementation manner of this specification, if the reproduction description information is the same as the abnormal description information, it indicates that the reproduction description information matches the abnormal description information. In another possible implementation manner of this specification, if the information similarity between the reproduction description information and the abnormal description information is greater than a preset threshold, it is determined that the reproduction description information matches the abnormal description information. Among them, the calculation methods of information similarity include but are not limited to the cosine similarity calculation method and the Euclidean distance calculation method. The preset threshold is specifically selected according to the actual situation, and the embodiments of this specification do not make any limitations in this regard.

[0105] Exemplarily, if the abnormal description information of the target abnormal data is "In the first sentence of the second paragraph on the first page, there is a problem of misspelling '发' as '法'", the obtained target reproduction data can be "Our company's development team has achieved remarkable results in the past year". If the abnormal description information is "On the user management platform, when a user attempts to change the password, the system always returns an error message [Password change failed, please try again]. The user reports that they have tried multiple times without success", the obtained target reproduction data can be "The user cannot change the password".

[0106] Applying the solution of the embodiments of this specification, after obtaining the reproduction data through reproduction, the reproduction effect of the reproduction data is judged using the reproduction description information, enabling the target reproduction data to successfully reproduce the abnormal problem of the target abnormal data, and improving the accuracy of abnormal data reproduction.

[0107] In an optional embodiment of this specification, a complete experience update operation is proposed, realizing the continuous learning and evolution of empirical knowledge. That is, after determining the reproduction data as the target reproduction data when the reproduction description information of the reproduction data matches the abnormal description information, the following steps may further be included:

[0108] Update the reproduction experience library according to the abnormal description information and the target reproduction data to obtain an updated reproduction experience library.

[0109] In practical applications, the reproduction experience library is updated according to the abnormal description information and the target reproduction data. There are many ways to obtain the updated reproduction experience library, which are selected according to the actual situation. The embodiments of this specification do not limit this. In one possible implementation of this specification, the abnormal description information and the target reproduction data can be input into the data processing model to obtain the target reproduction result and target reproduction experience of the target reproduction data, wherein the target reproduction result is used to describe the reproduction effect of the target reproduction data, and the target reproduction result includes reproduction success and reproduction failure; according to the target reproduction result and target reproduction experience, the reproduction experience library is updated to obtain the updated reproduction experience library. In another possible implementation of this specification, the reproduction experience, abnormal description information and target reproduction data in the reproduction experience library can be input into the data processing model to obtain the target reproduction result, target reproduction experience and target experience update suggestions of the target reproduction data, wherein the target experience update suggestions include at least one of adding suggestions, deleting suggestions, modifying suggestions and merging suggestions; according to the target experience update suggestions, target reproduction results and target reproduction experience, the reproduction experience library is updated to obtain the updated reproduction experience library.

[0110] By applying the solution of the embodiment of this specification, when the target reproduction data is determined, the reproduction experience library is updated according to the exception description information and the target reproduction data, an updated reproduction experience library is obtained, the reproduction experience is continuously accumulated and optimized, and the reproduction success rate is significantly improved.

[0111] In an optional embodiment of the present specification, the recurrence description information of the recurrence data may be obtained. That is, when the recurrence description information of the recurrence data matches the abnormal description information, before determining the recurrence data as the target recurrence data, the following steps may be further included:

[0112] Run the reproducible data and obtain the running log of the reproducible data;

[0113] Extract the recurrence description information of the recurrence data from the operation log, where the recurrence description information is used to describe the abnormal problems in the operation process of the recurrence data.

[0114] It should be noted that operation logs refer to log files generated during the operation of software or systems. These logs record various activities and status information of the reproduction data during operation. Operation logs are important for diagnosing problems, tracking errors, monitoring system performance, and auditing operations.

[0115] In practical applications, there are many ways to extract the recurrence description information of the recurrence data from the operation log, and the specific selection is based on the actual situation. The embodiments of this specification do not impose any restrictions on this. In one possible implementation of this specification, a pre-set description keyword can be obtained, and the operation log can be matched with the description keyword to obtain the recurrence description information of the recurrence data. In another possible implementation of this specification, the recurrence description information of the recurrence data can be extracted from the operation log through a log parsing tool or a parsing script.

[0116] By applying the solution of the embodiment of this specification, the reproduction description information of the reproduction data is extracted from the initial reproduction log, which ensures the accuracy of the reproduction description information. The reproduction description information can further be used to accurately determine whether the reproduction data successfully reproduces the abnormal problem.

[0117] In an optional embodiment of the present specification, a checking mechanism for abnormal data recurrence is also proposed, thereby improving the instructions for reproducing data, that is, the above-mentioned operation reproducing data, before obtaining the operation log of the reproducing data, may also include the following steps:

[0118] Obtaining a data inspection strategy, wherein the data inspection strategy includes at least one of a reproduction inspection strategy, an environment inspection strategy, and an operation inspection strategy;

[0119] Use the data inspection strategy to inspect the recurrence data and obtain the inspection results;

[0120] Run the recurrence data and obtain the running log of the recurrence data. The following steps can be included:

[0121] If the check result indicates that the running of the reproducing data is passed, run the reproducing data to obtain the running log of the reproducing data.

[0122] It should be noted that the data check strategy is a series of rules and methods developed to check the quality of the reproducing data. By using the data check strategy to check the reproducing data, you can obtain the check result of whether the reproducing data has errors. Specifically, the data check strategy can check whether the reproducing data truly reproduces the abnormal problem instead of simply outputting error information, verify the grammatical correctness of the reproducing data, check whether the reproducing data contains necessary import statements and environment settings, check whether the reproducing data covers the key steps in the abnormal description information, etc.

[0123] Data check strategies can be divided into reproduction check strategies, environment check strategies, and operation check strategies according to different inspection focuses. The reproduction check strategy is used to check whether the reproduction data truly reproduces the abnormal problem of the target abnormal data, rather than simply outputting error information. The environment check strategy is used to check whether the operating environment of the reproduction data is wrong. The operation check strategy is used to check whether the running instructions of the reproduction data are wrong. The inspection results include the reproduction data check passing and the reproduction data check failing. If the reproduction data check passes, it means that the reproduction data can be run and the operation log of the reproduction data can be obtained. If the reproduction data check fails, it means that the reproduction data cannot be run. At this time, the reproduction data can be automatically corrected to improve the quality of the reproduction data.

[0124] Next, taking the reproduction data as the reproduction code as an example, the reasons why the reproduction data check fails are explained with examples. According to statistics, 64% of the failures are due to reproduction code errors. The causes of the errors include but are not limited to reproduction target errors, code format errors, reproduction in a simulated manner, lack of necessary environment interaction components and configurations, and problems with logic or function usage. 13.33% of the failures are due to running instruction errors. The causes of the errors include but are not limited to running instructions that are different from the requirements in the target exception code or the regular running instruction methods of the code repository. 22.67% of the failures are due to environment dependency errors. The causes of the errors include but are not limited to the fact that the provided execution environment does not match the target exception code, and the code will not have errors in the target exception code after execution.

[0125] In practical applications, there are many ways to obtain data inspection strategies, which are selected according to actual conditions, and the embodiments of this specification do not limit this. In one possible implementation of this specification, a data inspection strategy sent by a user through a client can be received. In another possible implementation of this specification, the data inspection strategy can be read from other data acquisition devices or databases.

[0126] By applying the solution of the embodiment of this specification, the recurrence data is checked by using the data inspection strategy. When the inspection result indicates that the recurrence data inspection has passed, the recurrence data is run to obtain the operation log of the recurrence data, thereby ensuring the quality of the recurrence data.

[0127] In an optional embodiment of the present specification, after extracting the recurrence description information of the recurrence data from the operation log, the following steps may also be included:

[0128] When the reproduction description information of the reproduction data does not match the exception description information, return to execute the step of extracting the target reproduction experience corresponding to the target exception data from the reproduction experience library until the reproduction description information of the reproduction data matches the exception description information to obtain the target reproduction data.

[0129] It should be noted that if the reproduction description information of the reproduced data does not match the abnormal description information, it means that the reproduced data cannot successfully reproduce the abnormal problem in the target abnormal data. The reason for the failure of reproduction may be the lack of target reproduction experience. Therefore, the step of extracting the target reproduction experience corresponding to the target abnormal data from the reproduction experience library can be returned until the reproduction description information of the reproduced data matches the abnormal description information to obtain the target reproduction data. Among them, the target reproduction experience returned to be extracted can be different in experience content or different in experience quantity compared with the target reproduction experience previously used to reproduce the target abnormal data. The specific selection is made according to the actual situation and the embodiment of this specification does not impose any limitation on this.

[0130] Furthermore, when the reproduction description information of the reproduction data does not match the exception description information, return to execute the step of extracting the target reproduction experience corresponding to the target exception data from the reproduction experience library. If the preset number of iterations is reached and the reproduction description information of the reproduction data still does not match the exception description information, the reproduction is determined to have failed. The preset number of iterations is selected based on the actual situation, and the embodiments of this specification do not impose any restrictions on this.

[0131] By applying the solution of the embodiments of this specification, when the reproduction description information of the reproduction data does not match the abnormal description information, the target reproduction experience corresponding to the target abnormal data is re-extracted, thereby avoiding the reproduction failure caused by insufficient target reproduction experience and improving the success rate of abnormal data reproduction.

[0132] In an optional embodiment of the present specification, after extracting the recurrence description information of the recurrence data from the operation log, the following steps may also be included:

[0133] When the recurrence description information of the recurrence data does not match the abnormal description information, generating modification guidance information according to the recurrence description information and the abnormal description information;

[0134] Modify the recurrence data according to the modification guide information to obtain the modified recurrence data;

[0135] When the modification description information of the modified recurrence data matches the abnormal description information, the modified recurrence data is determined as the target recurrence data, wherein the modification description information is used to describe the abnormal problem during the operation of the modified recurrence data.

[0136] It should be noted that if the reproduction description information of the reproduction data does not match the abnormal description information, it means that the reproduction data cannot successfully reproduce the abnormal problem in the target abnormal data. The reproduction data can be modified and optimized according to the modification guide information to obtain the modified reproduction data. The modification guide information (prompt) is used to guide the process of modifying and optimizing the reproduction data. For example, the modification guide information can be "You are a professional programmer and hope that the reproduction data can reproduce the abnormal problem that matches the abnormal description information, but there is a gap between the reproduction descriptive information of the reproduction data and the abnormal description information. Please optimize the reproduction data and output the modified reproduction data." The modification guide information is selected according to the actual situation, and the embodiments of this specification do not impose any restrictions on this.

[0137] In actual applications, there are multiple ways to generate modification guidance information based on the recurrence description information and the exception description information, and the specific selection is based on the actual situation. The embodiments of this specification do not impose any restrictions on this. In one possible implementation of this specification, a modification guidance template can be obtained, and the recurrence description information and the exception description information can be added to the modification guidance template to obtain the modification guidance information. In another possible implementation of this specification, the difference information between the recurrence description information and the exception description information can be compared, and the difference information can be added to the modification guidance template to obtain the modification guidance information.

[0138] Furthermore, there are multiple ways to modify the recurring data according to the modification guide information to obtain the modified recurring data, which can be selected according to the actual situation, and the embodiments of this specification do not limit this. In one possible implementation of this specification, the data modification engine can be called to modify the recurring data according to the modification guide information to obtain the modified recurring data. In another possible implementation of this specification, the modification guide information and the recurring data can be input into the data processing model to obtain the modified recurring data.

[0139] By applying the solution of the embodiment of this specification, when the reproduction description information of the reproduction data does not match the abnormal description information, the reproduction data is optimized, thereby improving the success rate of abnormal data reproduction.

[0140] The following combination Figure 3 , taking the application of the abnormal data reproduction method provided in this specification in the software development scenario as an example, the abnormal data reproduction method is further explained. Figure 3 A flowchart of an abnormal code reproduction method provided by an embodiment of this specification is shown, which specifically includes the following steps:

[0141] Step 302: Obtain exception description information of the target exception code.

[0142] Step 304: extracting the target reproduction experience corresponding to the target abnormal code from the reproduction experience library, wherein the reproduction experience library includes a plurality of reproduction experiences, and the reproduction experience is used to guide the reproduction process of the abnormal code.

[0143] Step 306: Reproduce the target exception code according to the exception description information and the target reproduction experience to obtain the reproduction code.

[0144] It should be noted that the implementation of step 302 to step 306 is the same as the implementation of step 202 to step 206, and the embodiments of this specification do not impose any limitation on this.

[0145] By applying the solution of the embodiment of this specification, the abnormal code reproduction method can realize end-to-end automatic generation from abnormal description information to template reproduction code. Introducing target reproduction experience in the generation process improves the efficiency and success rate of abnormal code reproduction.

[0146] See also Figure 4 , Figure 4 A processing flow chart of an abnormal code reproduction method provided by an embodiment of the present specification is shown, obtaining the context code and abnormal description information of the target abnormal code; extracting the target reproduction experience corresponding to the target abnormal code from the reproduction experience library; inputting the context code, abnormal description information and target reproduction experience into the data processing model, and reproducing the target abnormal code by the data processing model to obtain the reproduction code;

[0147] Use the data inspection strategy to inspect the recurrence data and obtain the inspection result; if the inspection result indicates that the recurrence data inspection fails, optimize the recurrence code until the inspection result indicates that the recurrence data inspection passes; if the inspection result indicates that the recurrence data inspection passes, run the recurrence code and obtain the running log of the recurrence code;

[0148] Extract the reproduction description information of the reproduction code from the operation log, and use the reproduction description information and the exception description information to check whether the reproduction is successful: if the reproduction description information of the reproduction code does not match the exception description information, it means that the reproduction has failed, and the reproduction target exception code is returned, that is, the step of extracting the target reproduction experience corresponding to the target exception code from the reproduction experience library is executed until the reproduction description information of the reproduction code matches the exception description information, and the target reproduction code is obtained; if the reproduction description information of the reproduction code matches the exception description information, it means that the reproduction is successful, and the reproduction code is determined as the target reproduction code;

[0149] According to the exception description information and the target reproduction code, the reproduction experience library is updated to obtain an updated reproduction experience library.

[0150] By applying the solution of the embodiments of this specification, through the continuously evolving end-to-end warehouse abnormality problem reproduction solution proposed in the embodiments of this specification, the reproduction experience can be continuously accumulated and optimized. With the use of the reproduction experience, the system performance can be continuously improved, the reproduction success rate can be further significantly improved, manual intervention can be reduced, and the automatic generation of reproduction code from problem description can be realized.

[0151] After testing, the solution proposed in the embodiment of this specification improved the reproduction performance by 22% in the SWE-bench benchmark test, reaching a very advanced level (SOTA level), which can accurately locate abnormal problems and speed up the repair process, significantly improve the reproduction success rate, and provide important support for the realization of automatic software engineering and artificial intelligence programmers.

[0152] Corresponding to the above-mentioned abnormal data reproduction method embodiment, this specification also provides an abnormal data reproduction device embodiment, Figure 5 FIG. 1 is a schematic diagram showing the structure of an abnormal data reproduction device provided by an embodiment of the present specification. Figure 5 As shown, the device comprises:

[0153] A first acquisition module 502 is configured to acquire abnormal description information of target abnormal data;

[0154] The first extraction module 504 is configured to extract target recurrence experience corresponding to target abnormal data from a recurrence experience library, wherein the recurrence experience library includes a plurality of recurrence experiences, and the recurrence experience is used to guide the recurrence process of the abnormal data;

[0155] The first reproduction module 506 is configured to reproduce target abnormal data according to the abnormal description information and target reproduction experience to obtain reproduction data.

[0156] Optionally, the device further includes: a first determination module configured to determine the recurrence data as target recurrence data when the recurrence description information of the recurrence data matches the abnormal description information.

[0157] Optionally, the device also includes: an operation module, configured to run the reproduction data and obtain an operation log of the reproduction data; extract reproduction description information of the reproduction data from the operation log, wherein the reproduction description information is used to describe abnormal problems during the operation of the reproduction data.

[0158] Optionally, the device also includes: an inspection module, configured to obtain a data inspection strategy, wherein the data inspection strategy includes at least one of a reproduction inspection strategy, an environment inspection strategy, and an operation inspection strategy; using the data inspection strategy to inspect the reproduction data to obtain an inspection result; and an operation module, further configured to run the reproduction data and obtain an operation log of the reproduction data when the inspection result indicates that the reproduction data inspection has passed.

[0159] Optionally, the device also includes: an execution module, which is configured to return to execute the step of extracting target reproduction experience corresponding to target abnormal data from the reproduction experience library when the reproduction description information of the reproduction data does not match the abnormal description information, until the reproduction description information of the reproduction data matches the abnormal description information, and the target reproduction data is obtained.

[0160] Optionally, the device also includes: a generation module, configured to generate modification guidance information based on the reproduction description information and the exception description information when the reproduction description information of the reproduction data does not match the exception description information; modify the reproduction data according to the modification guidance information to obtain modified reproduction data; and determine the modified reproduction data as target reproduction data when the modification description information of the modified reproduction data matches the exception description information, wherein the modification description information is used to describe abnormal problems during the operation of the modified reproduction data.

[0161] Optionally, the device also includes: a construction module, configured to obtain description information of the abnormal data and reproduction data of the abnormal data; input the description information and the reproduction data into a data processing model to obtain reproduction results and reproduction experience of the reproduction data, wherein the reproduction results are used to describe the reproduction effect of the reproduction data; and construct a reproduction experience library based on the reproduction results and reproduction experience.

[0162] Optionally, the construction module is further configured to input the original reproduction experience, description information and reproduction data in the original experience library into the data processing model to obtain the reproduction results, reproduction experience and experience update suggestions of the reproduction data, wherein the experience update suggestions include at least one of adding suggestions, deleting suggestions, modifying suggestions and merging suggestions; based on the experience update suggestions, reproduction results and reproduction experience, the original experience library is updated to obtain the reproduction experience library.

[0163] Optionally, the device further includes: an updating module configured to update the reproduction experience library according to the exception description information and the target reproduction data to obtain an updated reproduction experience library.

[0164] Optionally, the device also includes: a third acquisition module, configured to obtain context data of target abnormal data; a first reproduction module 506, further configured to reproduce the target abnormal data according to the context data, abnormal description information and target reproduction experience to obtain reproduction data.

[0165] Optionally, the first extraction module 504 is further configured to obtain descriptive information corresponding to multiple reproduction experiences in the reproduction experience library; match the abnormal description information and the description information to obtain a matching result, wherein the matching result corresponds one-to-one to the description information; and based on the matching result, extract the target reproduction experience corresponding to the target abnormal data from the multiple reproduction experiences.

[0166] The solution of the embodiments of the present specification is applied to automatically reproduce the target abnormal data by introducing the target reproduction experience, thereby improving the efficiency and success rate of abnormal data reproduction. Moreover, after the reproduction data is obtained, the reproduction description information is used to judge the reproduction effect of the reproduction data, so that the target reproduction data can successfully reproduce the abnormal problems of the target abnormal data, thereby improving the accuracy of abnormal data reproduction.

[0167] The above is a schematic scheme of an abnormal data reproduction device of this embodiment. It should be noted that the technical scheme of the abnormal data reproduction device and the technical scheme of the abnormal data reproduction method described above belong to the same concept, and the details not described in detail in the technical scheme of the abnormal data reproduction device can be referred to the description of the technical scheme of the abnormal data reproduction method described above.

[0168] Corresponding to the above-mentioned abnormal code reproduction method embodiment, this specification also provides an abnormal code reproduction device embodiment, Figure 6 FIG. 1 is a schematic diagram showing the structure of an abnormal code reproduction device provided by an embodiment of the present specification. Figure 6 As shown, the device comprises:

[0169] The second acquisition module 602 is configured to acquire the exception description information of the target exception code;

[0170] The second extraction module 604 is configured to extract a target recurrence experience corresponding to a target abnormal code from a recurrence experience library, wherein the recurrence experience library includes a plurality of recurrence experiences, and the recurrence experience is used to guide the recurrence process of the abnormal code;

[0171] The second reproduction module 606 is configured to reproduce the target exception code according to the exception description information and the target reproduction experience to obtain the reproduction code.

[0172] By applying the scheme of the embodiment of this specification, through the abnormal code reproduction method, end-to-end automatic generation from abnormal description information to template reproduction code can be achieved. In the generation process, the target reproduction experience is introduced to improve the efficiency and success rate of abnormal code reproduction, and after the reproduction code is obtained, the reproduction effect of the reproduction code is judged by using the reproduction description information, so that the target reproduction code can successfully reproduce the abnormal problem of the target abnormal code, and the accuracy of abnormal code reproduction is improved, thereby improving the efficiency and accuracy of automatic data software engineering.

[0173] The above is a schematic scheme of an abnormal code reproduction device of this embodiment. It should be noted that the technical scheme of the abnormal code reproduction device and the technical scheme of the abnormal code reproduction method described above belong to the same concept, and the details not described in detail in the technical scheme of the abnormal code reproduction device can be referred to the description of the technical scheme of the abnormal code reproduction method described above.

[0174] Figure 7 A structural block diagram of a computing device 700 provided in one embodiment of the present specification is shown.

[0175] The computing device 700 includes:

[0176] Memory 710 and processor 720;

[0177] The memory 710 is used to store computer programs / instructions, and the processor 720 is used to execute the computer programs / instructions. When the computer programs / instructions are executed by the processor 720, the steps of the above-mentioned abnormal data reproduction method or abnormal code reproduction method are implemented.

[0178] In one or more embodiments of the present specification, the computing device may be understood as an integrated intelligent terminal, including but not limited to a server, a desktop computer, a PC (Personal Computer), an all-in-one model machine, a mobile phone, a tablet computer or other portable intelligent terminal, etc., and the computing device may be pre-installed with the model described in the above embodiments of the present application.

[0179] Specifically, the computing device can pre-set multiple types of models, including but not limited to models in the fields of natural language processing, visual processing, speech processing, code processing, multimodal task processing, etc., so as to provide a variety of model choices. In different product forms, the computing device can support one or more model usage methods, including but not limited to model training, model calling, model fine-tuning, model deployment, model reasoning and application, etc. In some product forms, the computing device also supports model management, including but not limited to multi-type model management (supporting the management of multiple types of models such as discriminants and generative models), model version control (supporting the control of different model versions), model evaluation (based on model evaluation tools, evaluating the performance and effect of the model), etc. In other product forms, the computing device can also create applications based on models, provide application programming interface (API, Application Programming Interface) calling capabilities, and can call models to the created applications through API interfaces, while providing application management tools to achieve management and monitoring of applications.

[0180] Furthermore, the computing device can also include data management (supporting the creation and management of model tuning data sets), a training center (providing rich training resources to help users learn and master artificial intelligence technology), and basic management and control capabilities (providing enterprise-level basic management and control capabilities to ensure the security and efficient operation of the system). Through the above functions, a comprehensive, integrated artificial intelligence development, training, deployment and application device is provided.

[0181] In addition, the components of the computing device 700 include, but are not limited to, a memory 710 and a processor 720. The processor 720 and the memory 710 may be connected via a bus.

[0182] The computing device 700 may also include an access device that enables the computing device 700 to communicate with a database storing data via one or more networks. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world microwave interconnection access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, and a near field communication (NFC).

[0183] In one embodiment of the present specification, the above components of the computing device 700 and Figure 7 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Figure 7 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0184] The computing device 700 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 700 may also be a mobile or stationary server.

[0185] The processor 720 is used to execute a computer program / instruction, which implements the steps of the above-mentioned abnormal data reproduction method or abnormal code reproduction method when executed by the processor.

[0186] The above is a schematic scheme of a computing device of this embodiment. It should be noted that the technical scheme of the computing device and the technical scheme of the above-mentioned abnormal data reproduction method and abnormal code reproduction method belong to the same concept, and the details not described in detail in the technical scheme of the computing device can be referred to the description of the technical scheme of the above-mentioned abnormal data reproduction method or abnormal code reproduction method.

[0187] An embodiment of the present specification further provides a computer-readable storage medium storing a computer program / instruction, which implements the steps of the above-mentioned abnormal data reproduction method or abnormal code reproduction method when executed by a processor.

[0188] The above is a schematic scheme of a computer-readable storage medium of this embodiment. It should be noted that the technical scheme of the storage medium and the technical scheme of the above-mentioned abnormal data reproduction method and abnormal code reproduction method belong to the same concept, and the details not described in detail in the technical scheme of the storage medium can be referred to the description of the technical scheme of the above-mentioned abnormal data reproduction method or abnormal code reproduction method.

[0189] An embodiment of the present specification also provides a computer program product, including a computer program / instruction, which implements the steps of the above-mentioned abnormal data reproduction method or abnormal code reproduction method when executed by a processor.

[0190] The above is a schematic scheme of a computer program product of this embodiment. It should be noted that the technical scheme of the computer program product and the technical scheme of the above-mentioned abnormal data reproduction method and abnormal code reproduction method belong to the same concept, and the details not described in detail in the technical scheme of the computer program product can be referred to the description of the technical scheme of the above-mentioned abnormal data reproduction method or abnormal code reproduction method.

[0191] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0192] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of patent practice. For example, in some regions, according to patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0193] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of this specification are not limited by the order of the actions described, because according to the embodiments of this specification, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of this specification.

[0194] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0195] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The optional embodiments do not describe all the details in detail, nor do they limit the invention to only the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of this specification, so that technicians in the relevant technical field can well understand and use this specification. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. A method for reproducing abnormal data, comprising: Get the abnormal description information of the target abnormal data; Extracting target recurrence experience corresponding to the target abnormal data from a recurrence experience database, wherein the recurrence experience database includes a plurality of recurrence experiences, and the recurrence experience is used to guide the recurrence process of the abnormal data; Reproduce the target abnormal data according to the abnormal description information and the target reproduction experience to obtain reproduction data.

2. The method according to claim 1, wherein the target abnormal data is reproduced according to the abnormal description information and the target reproduction experience, and after obtaining the reproduction data, the method further comprises: When the recurrence description information of the recurrence data matches the abnormal description information, the recurrence data is determined as the target recurrence data.

3. The method according to claim 2, wherein when the recurrence description information of the recurrence data matches the abnormal description information, before determining the recurrence data as the target recurrence data, further comprises: Run the reproduced data to obtain a running log of the reproduced data; Extract the recurrence description information of the recurrence data from the operation log, wherein the recurrence description information is used to describe abnormal problems in the operation process of the recurrence data.

4. The method according to claim 3, before running the recurring data and obtaining the running log of the recurring data, further comprises: Obtaining a data inspection strategy, wherein the data inspection strategy includes at least one of a reproduction inspection strategy, an environment inspection strategy, and an operation inspection strategy; Using the data inspection strategy to inspect the recurrence data, and obtaining an inspection result; The running of the reproducing data to obtain a running log of the reproducing data includes: When the inspection result indicates that the recurrence data inspection has passed, the recurrence data is run to obtain an operation log of the recurrence data.

5. The method according to claim 3, after extracting the recurrence description information of the recurrence data from the operation log, further comprises: In the case where the reproduction description information of the reproduction data does not match the exception description information, return to execute the step of extracting the target reproduction experience corresponding to the target exception data from the reproduction experience library until the reproduction description information of the reproduction data matches the exception description information to obtain the target reproduction data.

6. The method according to claim 3, after extracting the recurrence description information of the recurrence data from the operation log, further comprises: In a case where the recurrence description information of the recurrence data does not match the abnormal description information, generating modification guidance information according to the recurrence description information and the abnormal description information; Modify the recurring data according to the modification guide information to obtain modified recurring data; In a case where the modification description information of the modified and reproduced data matches the abnormal description information, the modified and reproduced data is determined as the target reproduction data, wherein the modification description information is used to describe the abnormal problem during the operation of the modified and reproduced data.

7. The method according to claim 1, before extracting the target reproduction experience corresponding to the target abnormal data from the reproduction experience library, further comprising: Obtaining description information of abnormal data and reproduction data of the abnormal data; Inputting the description information and the reproduction data of the abnormal data into a data processing model to obtain a reproduction result and a reproduction experience of the reproduction data, wherein the reproduction result is used to describe the reproduction effect of the reproduction data of the abnormal data; A reproduction experience library is constructed based on the reproduction result and the reproduction experience.

8. The method according to claim 7, wherein the step of inputting the description information and the reproduction data of the abnormal data into a data processing model to obtain the reproduction result and reproduction experience of the reproduction data comprises: Inputting the original recurrence experience in the original experience library, the description information, and the recurrence data of the abnormal data into the data processing model, obtaining the recurrence result of the recurrence data, the recurrence experience, and the experience update suggestion, wherein the experience update suggestion includes at least one of adding suggestion, deleting suggestion, modifying suggestion, and merging suggestion; The step of constructing a reproduction experience library according to the reproduction result and the reproduction experience includes: The original experience database is updated according to the experience update suggestion, the reproduction result and the reproduction experience to obtain a reproduction experience database.

9. The method according to claim 2, wherein after determining the recurrence data as target recurrence data when the recurrence description information of the recurrence data matches the exception description information, the method further comprises: The reproduction experience library is updated according to the exception description information and the target reproduction data to obtain an updated reproduction experience library.

10. The method according to claim 1, wherein the step of reproducing the target abnormal data according to the abnormal description information and the target reproduction experience, and obtaining the reproduction data, further comprises: Acquire context data of the target abnormal data; The reproducing the target abnormal data according to the abnormal description information and the target reproduction experience to obtain the reproduction data includes: The target abnormal data is reproduced according to the context data, the abnormal description information and the target reproduction experience to obtain reproduction data.

11. According to the method of claim 1, extracting the target reproduction experience corresponding to the target abnormal data from the reproduction experience library comprises: Obtaining description information corresponding to a plurality of recurrence experiences in the recurrence experience library; Matching the abnormal description information with the description information to obtain a matching result, wherein the matching result corresponds to the description information one by one; According to the matching result, a target reproduction experience corresponding to the target abnormal data is extracted from the multiple reproduction experiences.

12. A method for reproducing abnormal code, comprising: Get the exception description information of the target exception code; Extracting the target recurrence experience corresponding to the target abnormal code from the recurrence experience library, wherein the recurrence experience library includes a plurality of recurrence experiences, and the recurrence experience is used to guide the recurrence process of the abnormal code; According to the exception description information and the target reproduction experience, the target exception code is reproduced to obtain a reproduction code.

13. A computing device comprising: Memory and processor; The memory is used to store computer programs / instructions, and the processor is used to execute the computer programs / instructions. When the computer program / instructions are executed by the processor, the steps of the method described in any one of claims 1 to 11 or claim 12 are implemented.

14. A computer-readable storage medium storing a computer program / instruction, wherein the computer program / instruction, when executed by a processor, implements the steps of any one of claims 1 to 11 or the method of claim 12.

15. A computer program product, comprising a computer program / instruction, which, when executed by a processor, implements the steps of any one of claims 1 to 11 or the method of claim 12.

Images