Test strategy suggestion generation method and device based on problem distribution analysis, equipment and storage medium

Abstract

The invention discloses a test strategy suggestion generation method and device based on question distribution analysis, equipment and a storage medium, and relates to the field of statistical data processing, and the method comprises the steps: obtaining question information which comprises question description; training a preset model according to the problem description to obtain a target model; analyzing the problem information through the target model to obtain an analysis report; and generating a test strategy suggestion according to the analysis report. According to the invention, the accuracy and efficiency of problem distribution analysis can be improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method, apparatus, device, and storage medium for generating test strategy recommendations based on problem distribution analysis. Background Art

[0002] As the complexity of automotive software development projects continues to increase, project management tools like Jira are widely used to track and manage various issues encountered during projects. However, as project scale expands, the number of issues logged in Jira also increases dramatically. Efficiently analyzing the distribution of these issues, identifying the key modules and scenarios where they occur, and adjusting testing strategies accordingly has become a critical issue.

[0003] Traditional Jira issue analysis methods typically rely on manual statistics and analysis, which is not only time-consuming and labor-intensive but also prone to errors. With the rapid development of big model technology, it has become possible to use big models for data analysis. Summary of the Invention

[0004] The main purpose of this application is to provide a method, device, equipment and storage medium for generating test strategy recommendations based on problem distribution analysis, aiming to solve the technical problem of how to improve the accuracy and efficiency of problem distribution analysis.

[0005] To achieve the above objectives, this application proposes a method for generating test strategy suggestions based on problem distribution analysis, the method comprising:

[0006] Obtaining problem information, wherein the problem information includes a problem description;

[0007] Train the preset model according to the problem description to obtain the target model;

[0008] Analyze the problem information through the target model to obtain an analysis report;

[0009] Generate test strategy recommendations based on the analysis report.

[0010] In one embodiment, the step of training a preset model according to the question information to obtain a target model includes:

[0011] Select the model that meets the preset criteria as the preset model;

[0012] Extract key features of the problem description through a preset method;

[0013] The key features are input into the preset model for training to obtain a target model.

[0014] In one embodiment, the step of inputting the key features into the preset model for training to obtain a target model includes:

[0015] Initializing the preset model;

[0016] Define the loss function and choose an optimization algorithm;

[0017] The key features are input into the preset model, and the model parameters are updated through the loss function and the optimization algorithm to obtain the target model.

[0018] In one embodiment, the step of analyzing the problem information using the target model to obtain an analysis report includes:

[0019] Identifying module information and scenario information of the problem information through the target model;

[0020] Classify the questions according to the module information and the scenario information to obtain a classification result;

[0021] The classification results are analyzed to obtain an analysis report, which includes a module distribution map, a scenario distribution map, and a problem quantity statistics table.

[0022] In one embodiment, the step of generating a test strategy recommendation based on the classification report includes:

[0023] Based on the analysis report, identifying key modules and test scenarios for testing through the target model;

[0024] Generate test strategy recommendations based on the key modules and the test scenarios.

[0025] In one embodiment, the step of generating a test strategy recommendation based on the classification report includes:

[0026] After a preset time interval, return to the step of obtaining problem information and the step of updating the test strategy suggestion.

[0027] In one embodiment, the step of obtaining problem information includes:

[0028] Get problem data;

[0029] The problem data is preprocessed to obtain problem information.

[0030] In addition, to achieve the above-mentioned purpose, the present application also proposes a device for generating test strategy suggestions based on problem distribution analysis, the device comprising:

[0031] An acquisition module, configured to acquire problem information, wherein the problem information includes a problem description;

[0032] A training module is used to train the preset model according to the problem description to obtain a target model;

[0033] An analysis module, configured to analyze the problem information using the target model to obtain an analysis report;

[0034] A generating module is used to generate a test strategy suggestion according to the analysis report.

[0035] In addition, to achieve the above-mentioned purpose, the present application also proposes a test strategy recommendation generation device based on problem distribution analysis, the device including: a memory, a processor and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the test strategy recommendation generation method based on problem distribution analysis as described above.

[0036] In addition, to achieve the above-mentioned purpose, the present application also proposes a storage medium, which is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by the processor, the steps of the test strategy recommendation generation method based on problem distribution analysis as described above are implemented.

[0037] In addition, to achieve the above-mentioned purpose, the present application also provides a computer program product, which includes a computer program, and when the computer program is executed by a processor, it implements the steps of the test strategy recommendation generation method based on problem distribution analysis as described above.

[0038] One or more technical solutions proposed in this application have at least the following technical effects:

[0039] By automatically counting and analyzing recorded problems, identifying the modules and distribution of occurrence scenarios of the problems, outputting distribution statistics tables, and making recommendations for testing strategies, it solves the technical problems of traditional Jira problem analysis methods that usually rely on manual statistics and analysis, which is time-consuming, labor-intensive, and error-prone. Compared with existing technologies, it improves the accuracy and efficiency of problem distribution analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0041] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0042] Figure 1 A flowchart of the first embodiment of the method for generating test strategy recommendations based on problem distribution analysis provided in this application;

[0043] Figure 2 A flowchart of the second embodiment of the method for generating test strategy suggestions based on problem distribution analysis provided in this application;

[0044] Figure 3 A flowchart of the third embodiment of the method for generating test strategy recommendations based on problem distribution analysis of this application is provided;

[0045] Figure 4 A flowchart illustrating a fourth embodiment of a method for generating test strategy recommendations based on problem distribution analysis of this application is provided;

[0046] Figure 5 A flowchart of a fifth embodiment of a method for generating test strategy recommendations based on problem distribution analysis is provided for this application;

[0047] Figure 6 A schematic diagram of a simplified process of a method for generating test strategy recommendations based on problem distribution analysis provided in an embodiment of the present application;

[0048] Figure 7 This is a schematic diagram of the module structure of a test strategy suggestion generating device based on problem distribution analysis according to an embodiment of the present application;

[0049] Figure 8 Schematic diagram of the device structure of the hardware operating environment involved in the test strategy recommendation generation method based on problem distribution analysis in an embodiment of the present application.

[0050] The purpose, features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0051] It should be understood that the specific embodiments described herein are merely used to explain the technical solutions of the present application and are not intended to limit the present application.

[0052] In order to better understand the technical solution of the present application, a detailed description will be given below in conjunction with the accompanying drawings and specific implementation methods.

[0053] The main solution of the embodiment of the present application is: obtaining problem information, which includes a problem description; training a preset model according to the problem description to obtain a target model; analyzing the problem information through the target model to obtain an analysis report; and generating test strategy recommendations based on the analysis report.

[0054] In this embodiment, for ease of description, the following description is made with the internal executor of the test strategy suggestion generation system as the execution subject.

[0055] Because the existing Jira problem analysis method usually relies on manual statistics and analysis, it is time-consuming, labor-intensive and prone to errors.

[0056] This application provides a solution that uses automatic statistics and analysis of recorded problems, identifies the distribution of problem modules and occurrence scenarios, outputs distribution statistics tables, and makes recommendations on test strategies, thereby improving the accuracy and efficiency of problem distribution analysis.

[0057] As can be seen from the above embodiments, the present application adopts automatic statistics and analysis of recorded problems, identifies the distribution of problem modules and occurrence scenarios, outputs distribution statistics tables, and makes suggestions for testing strategies. It solves the technical problems that traditional Jira problem analysis methods usually rely on manual statistics and analysis, which are time-consuming, labor-intensive and error-prone, and achieves the improvement of the accuracy and efficiency of problem distribution analysis.

[0058] It should be noted that the execution entity of this embodiment can be a computing service device with data processing, network communication, and program execution capabilities, such as a tablet computer, personal computer, mobile phone, etc., or an electronic device capable of performing the aforementioned functions. This embodiment and the following embodiments are described below using the internal executor of the test strategy recommendation generation system as an example.

[0059] Based on this, the embodiment of the present application provides a method for generating test strategy suggestions based on problem distribution analysis, referring to Figure 1 , Figure 1 This is a flowchart of the first embodiment of the method for generating test strategy recommendations based on problem distribution analysis of this application.

[0060] In this embodiment, the method for generating test strategy suggestions based on problem distribution analysis includes steps S10 to S40:

[0061] Step S10: Obtain problem information, where the problem information includes a problem description.

[0062] It should be noted that the problem information is the problem summary, problem description, problem level, module label and other information extracted from the management tool; the problem description is a detailed text description of the matters that need to be tracked and handled (such as defects, tasks, requirements, etc.), including a title that can accurately summarize the core of the problem, so that people can roughly understand the problem at a glance, and at the same time cover detailed content, such as the precise steps to reproduce the problem, so that people can reproduce the problem according to the steps to better understand the situation; it also includes the scope of the impact of the problem on the project, system or business, such as a certain function cannot be used, data loss, performance degradation, etc.; the frequency of the problem is also included, such as it occurs every time, occasionally, or under specific conditions; relevant environmental information is also essential, including software version number, operating system type, browser category and version, etc., because different environments may cause different situations in which the problem occurs; the time record of the problem is discovered helps to trace the source of the problem; if there is additional information that can assist in explaining the problem, such as screenshots showing the error interface, log files providing detailed system records, error codes to help developers locate the problem, etc., they are also included in the problem description.

[0063] In addition, it should be noted that Jira is a powerful enterprise-level project management and issue tracking tool that provides a centralized platform for teams to plan, track and manage various project tasks, software defects, workflows and other matters. Through Jira, team members can clearly understand the progress of the project, the status of the problem and everyone's responsibilities. On this platform, workflows can be customized to adapt to different types of projects and team collaboration modes. For example, for software development projects, it can define the complete process from requirement collection, development, testing to deployment; for other non-software projects, the task flow can also be set according to actual business needs. Jira can support agile development methods such as Scrum and Kanban. In the Scrum mode, it can help manage sprint planning, task allocation and burn-down charts. In Kanban mode, it can well present the status of each task in the workflow (such as to-do, in progress, and completed), making it convenient for the team to intuitively grasp the progress of the work; it also provides a powerful permission management function to ensure that team members with different roles (such as project managers, developers, testers, stakeholders, etc.) can only access and operate the content within their responsibilities; and Jira can be integrated with a variety of other tools (such as code repositories, testing tools, etc.), allowing the entire project ecosystem to work more closely together and improve the overall efficiency of the project. This embodiment and the following embodiments are explained using the distribution analysis of Jira issues and test strategy recommendations as examples.

[0064] Step S20: training the preset model according to the problem description to obtain a target model.

[0065] It should be noted that the preset model is a large model; a large model, that is, a model with a large number of parameters trained through large-scale data and powerful computing power, is a combination of "big data + large computing power + strong algorithm", like an "implicit knowledge base" that condenses the inner essence of big data. The large model has significant characteristics. First, its parameters are huge, usually reaching billions or even trillions. Such a large number of parameters enables the model to capture extremely complex language patterns, knowledge relationships and various data features, and thus has powerful representation learning and complex task processing capabilities. Secondly, the scale of training data is also very considerable, requiring a large amount of text, images, voice and other types of data. These data come from a wide range of sources, covering various channels such as Internet pages, Wikipedia, books and papers, question-and-answer websites, etc. Moreover, the large model has a huge demand for computing power consumption. The training process requires the use of professional hardware equipment such as GPU, TPU, etc., and may last for weeks or even months, consuming a lot of time and electricity resources. In terms of working principles, the large model is first pre-trained through massive The model is initially trained on unlabeled data to allow it to learn the basic patterns and features in language, image and other data, build a basic language model or visual model, etc., so that it has preliminary generation capabilities, and then fine-tune. On the basis of pre-training, a small amount of labeled data is used to further optimize and adjust the model to better adapt to specific tasks or fields. When the number and complexity of the parameters of the large model reach a certain level, emergence will occur, that is, some new and unexpected capabilities will be generated. These capabilities are naturally generated in the process of interaction between a large number of parameters; the classification types of large models are rich and varied, and can be divided into language models, computer vision models and multimodal models according to data type. Language models specialize in processing and understanding human language, can read and generate text, and perform well in natural language processing tasks; computer vision models mainly process image and video data, enabling computers to "see" the content in images and videos; multimodal models can process and understand multiple types of data at the same time, and fuse and interact data from different modalities. According to the application field, it can be divided into general field big models, industry big models and vertical big models. General field big models have wide applicability and cross-domain generalization capabilities, and can handle various types of tasks; industry big models are trained according to the needs and data characteristics of specific industries, and can better meet the business needs of the industry; vertical big models focus on a specific vertical field, deeply explore the knowledge and laws of the field, and provide more accurate and in-depth analysis and answers for professional problems in specific fields.

[0066] By extracting features such as key words and phrases from the problem description and inputting them into the preset model for training, the preset model processes and learns these inputs based on the existing algorithm and initial parameters. During the learning process, the preset model continuously adjusts its parameters to better adapt to these input features, thereby improving its understanding and processing capabilities of the problem description. Through repeated training and parameter adjustment, the preset model gradually optimizes its performance. As training continues, the preset model's grasp of the characteristics of the problem description will become increasingly accurate, and its ability to handle problems will continue to improve. Ultimately, after sufficient training and optimization, the optimized preset model is obtained, which is the desired target model. It can process similar problem descriptions more efficiently and accurately, improve classification accuracy, and provide strong support for subsequent problem analysis and resolution.

[0067] Step S30: Analyze the problem information using the target model to obtain an analysis report.

[0068] A trained big model is used to comprehensively classify all Jira issues. Leveraging its powerful learning capabilities and extensive knowledge base, the big model accurately identifies the specific module and scenario for each Jira issue. During this process, the big model thoroughly analyzes each issue's description, relevant context, and possible correlation factors to accurately determine the module to which the issue belongs and the specific scenario in which it occurred. After classifying all Jira issues, the big model conducts further in-depth statistical analysis of the classification results. Using data analysis methods and efficient computing power, the classified issues are comprehensively sorted and summarized, generating a highly detailed report on the module and scenario distribution of the issues.

[0069] Step S40: Generate a test strategy recommendation based on the analysis report.

[0070] It should be noted that the test strategy recommendation is a comprehensive expression of a series of methods, plans and principles proposed in software testing or other types of testing activities in order to complete the testing tasks efficiently and comprehensively.

[0071] The Jira project management tool is currently commonly used in vehicle-mounted software development projects. This embodiment and the following embodiments are described using the application of test strategy recommendations to vehicle-mounted software development and testing as an example.

[0072] Based on the analysis report, comprehensive and in-depth testing strategy recommendations are generated that are targeted and effective. First, key information is extracted from the data and conclusions in the analysis report, such as problem clusters, frequently occurring defect types, performance bottlenecks, and potential security risks. For problem clusters, further analysis is conducted to determine the root causes, which may be design flaws in specific functional modules, improper handling of complex business logic, or problems with interaction with other modules. Based on this, the recommended testing strategy focuses on strengthening testing of these functional modules, increasing test case coverage, and employing a combination of testing methods, such as black-box and white-box testing, to more comprehensively detect potential issues. If the analysis report indicates frequent occurrence of specific defect types, such as vulnerabilities in data input validation, the recommended testing strategy should include a dedicated test plan for data input, with detailed equivalence class partitioning and boundary value analysis protocols. Comprehensive testing should be conducted across a wide range of possible input scenarios to ensure the software correctly handles both valid and invalid input data. For performance bottlenecks, a performance testing strategy should be developed based on the performance indicators identified in the report. Identify key performance indicators (KPIs), such as response time and throughput, and design performance test cases under different load scenarios. Gradually increase the system load and observe performance trends to promptly identify and resolve performance issues. Regarding security risks, develop a rigorous security testing strategy based on the security vulnerabilities revealed in the analysis report. Conduct comprehensive security vulnerability scans, utilizing professional security testing tools to detect potential security risks such as SQL injection and cross-site scripting attacks. Furthermore, strengthen permission management testing to ensure that different user roles can only access functions and data within their scope of authority, preventing unauthorized access and manipulation. Furthermore, consider other information in the analysis report, such as software usage scenarios and user behavior patterns, to develop a more realistic testing strategy. For example, if the software is primarily intended for in-vehicle use in a specific industry, consider the specific needs and usage conditions of that industry and conduct simulation tests for various possible in-vehicle conditions, including the impact of varying road conditions, temperature, humidity, and other factors on the software. Furthermore, allocate testing resources rationally, assigning experienced testers to key areas of testing based on the focus and difficulty of the testing strategy to ensure efficient testing. In addition, we continuously track and evaluate the execution effect of the test strategy, and adjust and optimize the test strategy in a timely manner according to the actual test situation to adapt to the ever-changing status and needs of the software.

[0073] This embodiment provides a method for generating test strategy recommendations based on problem distribution analysis. By automatically counting and analyzing recorded problems, identifying problem modules and the distribution of occurrence scenarios, outputting distribution statistics tables, and making recommendations for test strategies, it solves the technical problems of traditional Jira problem analysis methods that usually rely on manual statistics and analysis, which is time-consuming, labor-intensive, and error-prone, thereby improving the accuracy and efficiency of problem distribution analysis.

[0074] Based on the first embodiment of the present application, in the second embodiment of the present application, the same or similar contents as those in the above embodiment 1 can be referred to the above introduction and will not be described in detail later. Figure 2 , step S20 includes steps S21 to S23:

[0075] Step S21: Select a model that meets the preset standard as a preset model.

[0076] It should be noted that the preset standard model, that is, the preset model, is a large model suitable for text classification. There are various types of large models, and different types of models are suitable for different application scenarios. Large models suitable for text classification include BERT, GPT series, ERNIE series, T5, Electra, ChatGPT-type large models, etc. Among them, BERT performs well in NLP tasks such as text classification through pre-training and fine-tuning; the GPT series can convert classification problems into text generation tasks to determine text categories; the ERNIE series introduces prior knowledge such as knowledge graphs, which can better understand and process semantic information to improve classification accuracy; T5 regards all NLP tasks as text-to-text conversion problems to achieve text classification; Electra adopts a more efficient pre-training strategy, which can train and fine-tune faster to improve classification results; ChatGPT-type large models can perform text classification based on clue collection and reasoning. This embodiment and the following embodiments do not limit the specific selection of large models.

[0077] Step S22: extract key features of the problem description using a preset method.

[0078] It should be noted that the default method is natural language processing (NLP). NLP is a complex technical field that integrates knowledge from multiple disciplines, including computer science, artificial intelligence, and linguistics. It aims to enable computers to understand, generate, and process natural language used by humans, such as spoken and written text in daily communication. This technology uses complex algorithms and models to analyze natural language forms, such as text or speech, in various areas, including sentence structure analysis, lexical semantics recognition, and discourse coherence. Its applications are extensive, ranging from intelligent customer service systems that understand user inquiries and provide accurate responses, to machine translation software that converts between different languages, to text classification tools that classify large amounts of text into different categories. These applications, such as news classification and sentiment analysis, all rely on NLP. Furthermore, NLP can be used for automatic summarization to extract key content from text; in question-answering systems, to accurately answer various user questions; and even in text generation, to help create stories, poems, and other content. NLP acts as a bridge between computers and natural language, enabling computers to better understand human intent and respond appropriately.

[0079] Additionally, it's important to note that key features, such as keywords, phrases, and sentence structure, within a problem description play a crucial role in accurately conveying the essence of the problem, facilitating understanding of the problem context, and promoting effective problem resolution. Keywords, as key elements in a problem description, often provide a high-level overview of the core issues. For example, when describing a software failure, keywords like "crash," "no response," and "error code" can quickly convey the general nature of the problem to the recipient. These keywords act as problem labels, quickly guiding people to initially categorize and locate the problem. Phrases also play a crucial role in problem descriptions. They can more specifically address specific aspects or circumstances of the problem. For example, phrases like "network connection interrupted" and "file cannot be opened" clearly convey the specific symptoms of the problem. Phrases can provide more detailed clues for problem analysis, enabling problem solvers to conduct more targeted investigations and solutions. Sentence structure is also a key feature. Proper sentence structure can make problem descriptions clearer and more organized. For example, using the structure "When..., the problem..." clearly describes the conditions and symptoms of the problem. Good sentence structure helps avoid ambiguity and ensures accurate problem descriptions. Disorganized sentence structure can make the meaning of a question unclear, increasing the difficulty of problem solving. Key features of a question description, such as keywords, phrases, and sentence structure, together constitute a complete and accurate question expression. This not only helps the questioner better articulate the problem, but also provides important clues and evidence for the recipient, thereby improving the efficiency and quality of problem solving.

[0080] Step S23: input the key features into the preset model for training to obtain a target model.

[0081] Key features of the problem description, namely keywords, phrases, and sentence structures, are fed into the large model for in-depth training. These key features carry the core information and unique attributes of the problem, providing valuable learning material for the large model. Once these key features are fed into the large model, the model leverages its complex algorithms and powerful computing power to comprehensively analyze and process them, deeply exploring the connections between keywords, understanding the specific contexts expressed by phrases, and parsing the logical relationships inherent in sentence structures. During this process, the large model continuously adjusts its parameters to better adapt to the problem patterns represented by these key features. Through continuous training and optimization, the large model gradually improves its ability to understand and process problems, more accurately identifying the essence of problems and enabling more efficient classification and analysis of similar problems. Ultimately, the optimized large model, also known as the target model, is obtained. When presented with new problem descriptions, the target model leverages the data learned from the key features to more quickly and accurately process and respond, providing stronger support and assurance for problem resolution.

[0082] In a feasible implementation, step S23 includes steps S231 to S233:

[0083] Step S231: Initialize the preset model.

[0084] Initializing a large model first requires determining the model architecture and choosing an appropriate model type based on the specific task. For example, for natural language processing tasks, a Transformer architecture model may be considered, while for computer vision tasks, a convolutional neural network architecture may be more suitable. Model hyperparameters, including the number of layers, neurons, or heads, must also be defined. Hyperparameter selection can be determined by referring to rules of thumb, previous research, or through small-scale experiments. Next, data preparation is required: collecting data relevant to the model's application domain and task. For example, for language models, large amounts of text are needed, while for vision models, images or videos are needed. Once collected, data preprocessing is performed, including cleaning the data to remove noise, removing garbled text, and restoring pixels in images. The data must also be converted to a format acceptable to the model. For example, text must be segmented and mapped into word vectors, and images must be resized and pixel values ​​normalized. The data is then divided into training, validation, and test sets. Finally, initialize the model parameters and choose an appropriate initialization method, such as random initialization, sampling from a uniform distribution or normal distribution to initialize the weights; or use pre-trained model parameter initialization. If there is a pre-trained model for similar tasks, loading its parameters can speed up training and improve performance. Then initialize the parameters of each layer of the model according to the selected method.

[0085] From the perspective of model training, good initialization provides a reasonable starting point for model training. When using random initialization methods, appropriate initial values ​​are given to the model parameters. For example, in a neural network, weights are initialized according to a certain probability distribution (such as normal distribution or uniform distribution). This allows the model to break symmetry when training begins. If all parameters are initialized to the same value, then in the initial stage of training, different neurons or layers of the model may learn the same pattern and fail to effectively mine the complex features in the data. Through random initialization, each neuron has a different initial state, so that it can start learning the features in the data from different perspectives, laying the foundation for subsequent training.

[0086] Step S232: define the loss function and select the optimization algorithm.

[0087] In large models used for text classification, defining a loss function first requires clarifying that the classification task's goal is to minimize the difference between the predicted and true categories. A common loss function for classification problems is the cross-entropy loss function. Taking multi-class text classification as an example, assuming the model outputs a probability distribution for each category, the cross-entropy between the predicted probability distribution and the true label is calculated to measure the inaccuracy of the model's predictions. The cross-entropy loss function effectively penalizes any deviation between the model's predicted category probabilities and the actual category probabilities. When selecting an optimization algorithm, consider its efficiency and stability in updating parameters for large models. Stochastic gradient descent (SGD) is a basic optimization algorithm that updates parameters based on the gradient of a single example in each iteration. However, this approach can result in random parameter updates and slow convergence. To improve this, momentum methods can be used. Building on SGD, momentum takes into account the accumulation of previous gradients, giving parameter updates a certain amount of inertia and enabling faster convergence in the correct direction. Adaptive learning rate algorithms such as Adagrad, Adadelta, and RMSProp are also commonly used. These algorithms can automatically adjust the learning rate based on the historical gradient information of the parameters, so that each parameter has its own appropriate learning rate. This embodiment and the following embodiments do not limit the specific selection of loss function and optimization algorithm.

[0088] Step S233: input the key features into the preset model, update the model parameters through the loss function and the optimization algorithm, and obtain the target model.

[0089] Inputting key features into the large model and updating model parameters using a loss function and optimization algorithm is a tightly coupled, cyclical process. First, text data is fed into the large model. The model's forward propagation calculates the predicted probability distribution for each category. Then, using a defined loss function, such as the cross-entropy loss function, the loss between the predicted probabilities and the true labels is calculated. This loss represents the degree of error in the model's predictions. Next, using a selected optimization algorithm, such as the Adam optimization algorithm, the gradient of the loss function with respect to the model parameters is calculated. This critical step is crucial. The gradient indicates the rate of change of the loss function in the direction of each parameter, guiding us as to how to adjust the parameters to minimize the loss. When calculating the gradient, the optimization algorithm considers previously accumulated gradient information or adaptively adjusts the learning rate to more accurately estimate the direction and magnitude of the parameter update. The optimization algorithm then updates the model parameters based on the calculated gradient. The update proceeds in the opposite direction of the gradient. Since a positive gradient increases the loss function, updating in the opposite direction can reduce the loss. The magnitude of the update is determined by the learning rate, which controls the step size of each update. As new text data is continuously input into the model, the above process of forward propagation, loss calculation, gradient calculation and parameter update is repeated. The model parameters are gradually optimized and the loss value gradually decreases until the model converges, that is, the loss value no longer decreases significantly or reaches the pre-set stopping condition. Finally, the optimized large model, that is, the target model, is obtained.

[0090] By training large models, we can better classify text and improve the accuracy of classification.

[0091] This embodiment provides a method for generating test strategy recommendations based on problem distribution analysis, selecting a model that meets preset standards as a preset model; extracting key features of the problem description through a preset method; inputting the key features into the preset model for training to obtain a target model, thereby improving classification accuracy.

[0092] Based on the first embodiment of the present application, in the third embodiment of the present application, the same or similar contents as those in the above embodiment 1 can be referred to the above introduction and will not be described in detail later. Figure 3 , step S30 includes steps S31 to S33:

[0093] Step S31: Identify module information and scenario information of the problem information through the target model.

[0094] It should be noted that module information refers to the information of the module to which the problem belongs. Specifically, it refers to a specific part in a complex system, project or knowledge system that is divided according to factors such as the nature of the problem, the functional scope involved, and the related business field. It is used to classify problems and facilitate the location, analysis and resolution of problems. For example, in a software system, it may include user interface modules, data storage modules, business logic modules, etc. When the software cannot correctly save the data entered by the user, it can be determined that the problem belongs to the data storage module; in an enterprise's management system, there will be financial modules, human resources modules, marketing modules, etc. If the problem is about errors in the calculation of employee salaries, then it belongs to the human resources module. It helps to focus on the direction of problem resolution and allows relevant personnel to quickly find the corresponding professional knowledge, technical means or responsible team to handle the problem.

[0095] In addition, it should be noted that scenario information is a comprehensive description of multiple factors such as the specific environment, relevant conditions, and background circumstances when the problem occurred. It covers content such as the time and place where the problem occurred, the operations being performed at the time, the status of the equipment or system involved, the surrounding conditions related to it, and the relevant behavioral characteristics of the users. For example, in the software usage scenario, a problem occurs when a certain software is opened at a specific time on a specific version of the operating system to perform a specific function. The operating system version, the specific operation process, and the software interface display at the time constitute the scenario information of the problem. For example, in a daily life scenario, if a home appliance malfunctions, the weather at the time, whether it is frequently used, and whether there are other electrical appliances in the surrounding area interfering with it also constitute the scenario information of the problem. These help to fully understand the ins and outs of the problem, and thus better analyze and solve the problem.

[0096] Step S32: classify the questions according to the module information and the scenario information to obtain classification results.

[0097] First, the question text containing module information and scenario information is input into the big model. The big model uses the semantic understanding and knowledge representation capabilities learned during the training phase to conduct an in-depth analysis of the question text, extracting module-related keywords in the text, such as "user interface" and "database" in the software system, as well as keywords related to scenario information, such as "high-concurrency operation" and "operation on specific devices". These keywords are then matched with the knowledge graph or classification rules built within the model. The model evaluates the probability distribution of problems in different categories by comprehensively considering module information and scenario information. For example, for a software problem, the model will determine whether it belongs to categories such as functional module failure, performance problem or compatibility problem, while considering whether the scenario occurs in a normal user operation scenario, a high-load operation scenario or a special configuration scenario. Ultimately, a problem classification that best meets the module and scenario characteristics is determined and the classification result is output.

[0098] Step S33: Analyze the classification results to obtain an analysis report, which includes a module distribution map, a scenario distribution map, and a problem quantity statistics table.

[0099] The classification results include information about the module to which the problem belongs and the scenario in which it occurs. The information in the classification results is analyzed to form an intuitive chart to represent the distribution of the problem and other situations.

[0100] The module distribution map can intuitively display the distribution of issues in each module. Different modules can be distinguished by different colors, shapes, or marks, allowing users to see at a glance which modules have more concentrated issues and which modules have relatively fewer issues. The scenario distribution map can clearly show the frequency and distribution characteristics of issues in different scenarios, helping users understand the patterns of problem occurrence in different scenarios. The problem quantity statistics table lists the number of issues in each module and scenario in detail, presenting the distribution of issues in specific numerical form, providing users with accurate data reference. Through this comprehensive report, users can gain an in-depth understanding of the overall distribution of Jira issues, providing a strong basis for subsequent problem solving, project management, and decision making.

[0101] This embodiment provides a method for generating test strategy recommendations based on problem distribution analysis, which identifies the module information and scenario information of the problem information through the target model; classifies the problems according to the module information and the scenario information to obtain classification results; analyzes the classification results to obtain an analysis report, which includes a module distribution map, a scenario distribution map, and a problem quantity statistics table, thereby improving the efficiency of problem distribution analysis.

[0102] Based on the first embodiment of the present application, in the fourth embodiment of the present application, the same or similar contents as those in the above-mentioned embodiment 1 can be referred to the above introduction and will not be described in detail later. Figure 4 , step S40 includes steps S41 to S42:

[0103] Step S41 : Based on the analysis report, identify key modules and test scenarios for testing through the target model.

[0104] The relevant data from the module distribution map, scenario distribution map, and problem statistics table in the analysis report is organized and input into the big model in a suitable text format. Leveraging its powerful semantic understanding and feature extraction capabilities, the big model conducts in-depth analysis of this input. From the module distribution map, the big model identifies key factors such as the proportion of each module in the overall system, its importance, and the inter-module relationships. Combined with the frequency of problems reported by the corresponding modules in the problem statistics table, modules with a high number of problems and a potentially greater impact on the overall system are prioritized for testing. For example, if a business logic processing module frequently encounters a high number of problems in the problem statistics table, the big model identifies it as a key testing module. For the scenario distribution map, the big model analyzes the probability and complexity of different scenarios, as well as their interactions with other modules. By comparing the number of problems reported in each scenario in the problem statistics table, the big model identifies scenarios with high problem frequency, complexity, or significant interaction with key modules as key testing scenarios. For example, if a high-concurrency business operation scenario has a significant number of issues and is closely related to multiple key modules, the big model will identify that scenario as one requiring focused testing. Through this comprehensive analysis and judgment of multiple aspects of information, the big model ultimately accurately identifies the key modules and test scenarios for testing, providing a strong basis for the subsequent development of more targeted and efficient testing strategies.

[0105] Step S42: Generate test strategy recommendations based on the key modules and the test scenarios.

[0106] In the context of in-vehicle software development, key modules, such as the in-vehicle entertainment system, undergo comprehensive functional testing, covering issues such as audio playback compatibility with various formats, video playback smoothness and format support, Bluetooth connection stability and compatibility with different mobile phones, and radio channel search accuracy and signal reception strength. Performance testing also measures system startup speed, audio and video loading speed, and responsiveness and resource usage during multi-tasking. Long-term continuous playback testing can also be conducted to ensure stability. Security testing examines malware protection and user permission management to prevent unauthorized access and ensure secure data handling during power outages. For the navigation system module, functional testing verifies map accuracy, route planning rationality, the functionality of different navigation modes, real-time traffic information updates and display, and comprehensive destination search. Performance testing assesses response time and map data download and update efficiency. Security testing ensures immunity to malicious route tampering, robustness against signal interference, and compatibility with other in-vehicle systems. Functional testing of vehicle control modules encompasses basic control functions such as start-stop, acceleration, and deceleration, advanced features such as cruise control and automatic parking, and vehicle status monitoring accuracy. Performance testing evaluates control command response time and stability in extreme environments. Security testing prevents hacker attacks and illegal control and tests fault diagnosis and fault tolerance. For test scenarios such as high-temperature environments, the vehicle system is placed in high temperatures to simulate a temperature rise inside the vehicle. The system's stability and performance at high temperatures, including the screen display and the operating status of electronic components, are tested. Similarly, low-temperature environment scenarios test the system's performance at low temperatures. High-altitude scenarios consider the impact of air pressure changes on the system and test its functionality and performance at different altitudes. Driving vibration scenarios simulate the vibrations of a driving vehicle and test the system's reliability and connection stability under vibration. Multi-device connection scenarios test the compatibility and performance of multiple devices when connected to the vehicle system simultaneously. Through these comprehensive testing strategies, the stability, security, and functionality of the vehicle software are ensured in various situations.

[0107] In a feasible implementation manner, step S40 includes step S01:

[0108] Step S01, after a preset time interval, returns to the step of obtaining problem information and updating the step of recommending a test strategy.

[0109] It should be noted that the preset time is the update interval determined based on actual test conditions and problem data, that is, the update frequency.

[0110] After the update frequency is reached, return to the step of obtaining problem information and updating the test strategy recommendation step. That is, the big model regularly extracts new problem data and status updates of existing problems from Jira, and updates the results of the problem distribution analysis and the test strategy recommendation report.

[0111] The big model regularly extracts new issue data and status updates from Jira, enabling continuous optimization of issue distribution analysis results and test strategy recommendations. It also dynamically adjusts and optimizes test strategies based on the resolution of issues and the effectiveness of test strategy execution. If certain issues are effectively resolved using specific test strategies, the big model can summarize the experience and apply the successful testing methods to other similar issues. If certain test strategies are ineffective, the big model can analyze the causes and provide improvement suggestions, continuously refining the test strategy to adapt to the changes and challenges in the automotive software development process.

[0112] By setting the update frequency and updating the problem distribution analysis data, the accuracy and real-time nature of the model data are maintained.

[0113] This embodiment provides a method for generating test strategy recommendations based on problem distribution analysis. Based on the analysis report, the key modules and test scenarios of the test are identified through the target model; based on the key modules and the test scenarios, test strategy recommendations are generated, thereby maintaining the accuracy and real-time performance of the model data.

[0114] Based on the first embodiment of the present application, in the fifth embodiment of the present application, the same or similar contents as those in the above embodiment 1 can be referred to the above introduction and will not be described in detail later. Figure 5 , step S10 includes steps S11 to S12:

[0115] Step S11, obtaining question data.

[0116] It should be noted that problem data is unprocessed problem information, that is, the original state of the problem information in Jira, including problem summary, problem description, problem level, module label and other information.

[0117] Step S12: pre-process the problem data to obtain problem information.

[0118] The text descriptions of the problem data are cleaned to remove typos, null values, invalid information, unnecessary punctuation, meaningless special characters, and repeated expressions, ensuring a clearer and more accurate description of the problem phenomenon. Next, data with varying formats, such as time and severity, are standardized. Time recording formats are standardized, and severity levels are categorized according to established standards. Information such as the module to which the problem belongs and the scenario in which it occurred are accurately labeled. Any errors or ambiguities are corrected and clarified based on relevant background knowledge and further analysis of the problem. Missing values ​​in the data are also handled appropriately. If key information is missing, it can be supplemented by reviewing relevant records or interviewing relevant personnel. If this is impossible, reasonable inferences can be made based on the overall data characteristics and common problem scenarios, or the data is marked as missing. The processed data is then consolidated according to appropriate logical relationships to facilitate subsequent analysis, statistics, and modeling. For example, data from problems with different statuses within the same module can be organized together. This completes the preprocessing of the problem data and yields problem information.

[0119] This embodiment provides a method for generating test strategy recommendations based on problem distribution analysis, which obtains problem data; preprocesses the problem data to obtain problem information, thereby improving the quality and availability of information and data.

[0120] For example, to help understand the implementation process of the test strategy suggestion generation method based on problem distribution analysis obtained by combining the first, second, third, fourth and fifth embodiments, please refer to Figure 6 , Figure 6 A brief flowchart of a test strategy recommendation generation method based on problem distribution analysis is provided, specifically: S1, data collection and preprocessing; S2, large model training; S3, problem distribution analysis; S4, test strategy generation; S5, model data update and maintenance.

[0121] It should be noted that the above examples are only used to understand this application and do not constitute a limitation on the test strategy recommendation generation method based on problem distribution analysis of this application. More simple transformations based on this technical concept are all within the scope of protection of this application.

[0122] This application also provides a test strategy suggestion generation device based on problem distribution analysis, please refer to Figure 7 The device comprises:

[0123] The acquisition module 10 is used to acquire problem information, where the problem information includes a problem description.

[0124] The training module 20 is used to train the preset model according to the problem description to obtain a target model.

[0125] The analysis module 30 is used to analyze the problem information through the target model to obtain an analysis report.

[0126] The generating module 40 is configured to generate a test strategy suggestion based on the analysis report.

[0127] In one embodiment, the training module 20 is further used to select a model that meets preset standards as a preset model; extract key features of the problem description through a preset method; input the key features into the preset model for training to obtain a target model.

[0128] In one embodiment, the training module 20 is further used to initialize the preset model; define a loss function and select an optimization algorithm; input the key features into the preset model, update the model parameters through the loss function and the optimization algorithm, and obtain the target model.

[0129] In one embodiment, the analysis module 30 is also used to identify the module information and scenario information of the problem information through the target model; classify the problems according to the module information and the scenario information to obtain classification results; analyze the classification results to obtain an analysis report, and the analysis report includes a module distribution map, a scenario distribution map, and a problem quantity statistics table.

[0130] In one embodiment, the generating module 40 is further configured to identify key modules and test scenarios of the test through the target model based on the analysis report; and generate test strategy recommendations according to the key modules and the test scenarios.

[0131] In one embodiment, the generating module 40 is further configured to return to the step of obtaining problem information and updating the step of recommending a test strategy after a preset time interval.

[0132] In one embodiment, the acquisition module 10 is further configured to acquire problem data and pre-process the problem data to obtain problem information.

[0133] The test strategy suggestion generation device based on problem distribution analysis provided by the present application adopts the test strategy suggestion generation method based on problem distribution analysis in the above-mentioned embodiment, which can solve the technical problem of how to improve the accuracy and efficiency of problem distribution analysis. Compared with the prior art, the beneficial effects of the test strategy suggestion generation device based on problem distribution analysis provided by the present application are the same as the beneficial effects of the test strategy suggestion generation method based on problem distribution analysis provided by the above-mentioned embodiment, and the other technical features of the test strategy suggestion generation device based on problem distribution analysis are the same as the features disclosed in the above-mentioned embodiment method, which will not be repeated here.

[0134] The present application provides a device for generating test strategy recommendations based on problem distribution analysis. The device for generating test strategy recommendations based on problem distribution analysis includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for generating test strategy recommendations based on problem distribution analysis in the above-mentioned embodiment one.

[0135] Reference below Figure 8 , which shows a schematic structural diagram of a device for generating test strategy suggestions based on problem distribution analysis suitable for implementing an embodiment of the present application. The device for generating test strategy suggestions based on problem distribution analysis in the embodiment of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 8 The test strategy suggestion generating device based on problem distribution analysis shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0136] like Figure 8As shown, the test strategy suggestion generation device based on problem distribution analysis may include a processing device 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. Various programs and data required for the operation of the test strategy suggestion generation device based on problem distribution analysis are also stored in RAM 1004. The processing device 1001, ROM 1002, and RAM 1004 are connected to each other via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 1009. The communication device 1009 can allow the test strategy suggestion generation device based on problem distribution analysis to communicate wirelessly or wired with other devices to exchange data. Although the figure shows a test strategy suggestion generation device based on problem distribution analysis with various systems, it should be understood that it is not required to implement or have all the systems shown. More or fewer systems can be implemented or have instead.

[0137] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program comprising program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by the processing device 1001, the above-mentioned functions defined in the method of the embodiment disclosed in the present application are executed.

[0138] The test strategy suggestion generation device based on problem distribution analysis provided by this application adopts the test strategy suggestion generation method based on problem distribution analysis in the above-mentioned embodiment, which can solve the technical problem of how to improve the accuracy and efficiency of problem distribution analysis. Compared with the prior art, the beneficial effects of the test strategy suggestion generation device based on problem distribution analysis provided by this application are the same as the beneficial effects of the test strategy suggestion generation method based on problem distribution analysis provided by the above-mentioned embodiment, and the other technical features of the test strategy suggestion generation device based on problem distribution analysis are the same as the features disclosed in the method of the previous embodiment, which will not be repeated here.

[0139] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0140] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0141] The present application provides a computer-readable storage medium having computer-readable program instructions (ie, a computer program) stored thereon, and the computer-readable program instructions are used to execute the test strategy recommendation generation method based on problem distribution analysis in the above-mentioned embodiment.

[0142] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, systems or devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, system or device. The program code contained on the computer-readable storage medium may be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0143] The computer-readable storage medium may be included in the device for generating test strategy suggestions based on problem distribution analysis; or may exist independently without being assembled into the device for generating test strategy suggestions based on problem distribution analysis.

[0144] The above-mentioned computer-readable storage medium carries one or more programs. When the above-mentioned one or more programs are executed by a test strategy recommendation generation device based on problem distribution analysis, the test strategy recommendation generation device based on problem distribution analysis: obtains problem information, and the problem information includes a problem description; trains a preset model according to the problem description to obtain a target model; analyzes the problem information through the target model to obtain an analysis report; and generates a test strategy recommendation according to the analysis report.

[0145] Computer program code for performing the operations of the present application may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on the remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0146] The flow charts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present application. In this regard, each box in the flow chart or block diagram can represent a module, program segment or a part of code, and the module, program segment or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented by a dedicated hardware-based system that performs the specified function or operation, or can be implemented by a combination of dedicated hardware and computer instructions.

[0147] The modules described in the embodiments of the present application may be implemented in software or hardware, wherein the name of a module does not necessarily limit the unit itself.

[0148] The readable storage medium provided by this application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., a computer program) for executing the above-mentioned method for generating test strategy recommendations based on problem distribution analysis, and can solve the technical problem of how to improve the accuracy and efficiency of problem distribution analysis. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by this application are the same as the beneficial effects of the method for generating test strategy recommendations based on problem distribution analysis provided by the above-mentioned embodiment, and will not be elaborated here.

[0149] The present application also provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for generating test strategy suggestions based on problem distribution analysis.

[0150] The computer program product provided in this application can solve the technical problem of improving the accuracy and efficiency of problem distribution analysis. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the method for generating test strategy recommendations based on problem distribution analysis provided in the above-mentioned embodiment, and will not be elaborated here.

[0151] The above description is only part of the embodiments of the present application and does not limit the patent scope of the present application. All equivalent structural transformations made by using the contents of the present application specification and drawings under the technical concept of the present application, or direct / indirect application in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for generating test strategy suggestions based on problem distribution analysis, characterized in that: The method comprises: Obtaining problem information, wherein the problem information includes a problem description; The preset model is trained according to the problem description to obtain a target model; Analyze the problem information through the target model to obtain an analysis report; A test strategy recommendation is generated based on the analysis report.

2. The method according to claim 1, characterized in that The step of training the preset model according to the problem information to obtain the target model comprises: Select a model that meets the preset criteria as the preset model; Extract key features of the problem description by a preset method; The key features are input into the preset model for training to obtain a target model.

3. The method according to claim 2, characterized in that The step of inputting the key features into the preset model for training to obtain the target model comprises: Initializing the preset model; Define the loss function and choose an optimization algorithm; The key features are input into the preset model, and the model parameters are updated through the loss function and the optimization algorithm to obtain the target model.

4. The method according to claim 1, characterized in that The step of analyzing the problem information through the target model to obtain an analysis report comprises: Identify module information and scenario information of the problem information through the target model; Classify the questions according to the module information and the scenario information to obtain classification results; The classification results are analyzed to obtain an analysis report, which includes a module distribution map, a scenario distribution map, and a problem quantity statistics table.

5. The method according to claim 1, characterized in that The step of generating a test strategy suggestion according to the classification report comprises: Based on the analysis report, identifying key modules and test scenarios of the test through the target model; Generate test strategy recommendations based on the key modules and the test scenarios.

6. The method according to claim 1, characterized in that The step of generating a test strategy suggestion according to the classification report includes: After a preset time interval, return to the step of obtaining the problem information and update the step of recommending the test strategy.

7. The method according to claim 1, characterized in that The step of obtaining problem information comprises: Get the problem data; The problem data is preprocessed to obtain problem information.

8. A device for generating test strategy suggestions based on problem distribution analysis, characterized in that: The device comprises: An acquisition module, used to acquire problem information, wherein the problem information includes a problem description; A training module, used to train a preset model according to the problem description to obtain a target model; An analysis module, used to analyze the problem information through the target model to obtain an analysis report; A generation module is used to generate a test strategy suggestion according to the analysis report.

9. A device for generating test strategy suggestions based on problem distribution analysis, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the method for generating test strategy suggestions based on problem distribution analysis as described in any one of claims 1 to 7.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the method for generating test strategy recommendations based on problem distribution analysis as described in any one of claims 1 to 7 are implemented.

Images