Test method and system for low-altitude service platform based on AI new model

Abstract

The invention relates to a test method and system for a low-altitude service platform based on an AI new model, and the method comprises the steps: collecting test task data related to the low-altitude service platform, including user behaviors, service logic and historical test data, and carrying out the preprocessing and marking of the test task data; performing full-parameter fine tuning on the large language model by using the data to generate a test case model and a test script model, and deploying the test case model and the test script model in a test system; after a tester uploads a function screenshot and a related text, calling and triggering a test case model for analysis through an event-driven mechanism and an API (Application Program Interface), automatically generating a preliminary test case, and adjusting, storing and updating the preliminary test case according to actual requirements; and the test script model analyzes the adjusted test case, generates an automatic test script, performs automatic test and records a result, and finally generates a test report. According to the method and the device, automatic generation and adjustment of the test cases and the test scripts are realized, and the visualization and manageability of the test are improved.

Description

Technical Field

[0001] The present invention relates to the field of automated testing technology, and mainly to a testing method and system for a low-altitude business platform based on a new AI model. Background Art

[0002] With the rapid development of Internet technology, the business logic and user behavior involved in low-altitude business platforms have become increasingly complex. Traditional testing technology mainly relies on manually written test case scripts, which leads to generally low testing efficiency, insufficient test coverage, and long test time. Traditional testing methods are difficult to fully cover all key scenarios, so potential problems may be overlooked. Therefore, the use of more efficient testing methods, such as automated testing, intelligent scenario simulation, and test analysis based on big data, has become the key to improving test quality and efficiency. These advanced technologies can more effectively respond to complex business needs, improve test coverage, and shorten test cycles, thereby ensuring the stability and reliability of low-altitude business platforms in various scenarios.

[0003] For example, WO2018010552A1 "Testing method and apparatus" discloses "a testing method and apparatus, the method comprising: obtaining a test case consisting of one or more test templates, wherein the test template includes one or more test steps, the test steps are operations performed on the test object, and part or all of the one or more test steps are described in natural language; translating the test steps in the test case into machine language through the correspondence between natural language and machine language, wherein the machine language is a language that can be recognized and executed by the machine; executing the test case on the test object using the machine language; and recording the test results. Using natural language to write test cases and creating test cases through pre-edited templates saves the trouble of writing test cases because the test cases themselves can be directly executed. The invention reduces the steps of script development in the process of automated testing and solves the problem in the prior art that the use of script compilation for automated testing requires testers to have programming skills". However, the invention relies on pre-edited test templates, which can contain test steps described in natural language. The flexibility of this method is limited, especially when facing complex scenarios or test tasks that require dynamic adjustment. Templated test cases may not be able to fully meet the needs. Although the invention reduces manual script development through the translation of natural language and machine language, its test case generation is still based on pre-defined templates and natural language parsing. The parsing process may have certain limitations, and the efficiency of generating test cases depends on the completeness of the template and the accuracy of the translation mechanism, and the efficiency of automated test generation is low.

[0004] In order to solve the above-mentioned shortcomings, there is an urgent need for an automated testing method that reduces manual intervention and greatly improves testing efficiency. Summary of the invention

[0005] In order to solve the above-mentioned problems existing in the prior art, the present application provides a testing method and system for low-altitude business platforms based on a new AI model.

[0006] The technical solution of this application is as follows:

[0007] A testing method for a low-altitude business platform based on a new AI model, the method comprising:

[0008] Collect test task data related to the low-altitude business platform, including user behavior, business logic and historical test data; use the test task data to fine-tune the large language model, obtain the test case model and test script model, and deploy them in the test system;

[0009] The tester uploads the functional screenshots and related texts corresponding to the low-altitude business platform in the input module of the test system, analyzes the functional screenshots and related texts through the event-driven mechanism and API call triggering test case model, and obtains the case analysis results;

[0010] Automatically generate preliminary test cases based on the use case analysis results; testers adjust the test cases based on actual needs and automatically save and update;

[0011] Based on the event-driven mechanism and API calls, the test cases are automatically saved and updated to trigger the test script model to analyze them and obtain the script analysis results;

[0012] Generate an automated test script using the request test framework and Python programming language based on the script analysis results, and the tester adjusts the automated test script according to actual needs and saves and updates it;

[0013] Run the adjusted automated test script to perform automated testing, capture and record the corresponding test results, generate and store a test report based on the corresponding test results.

[0014] Preferably, the test task data is collected from the low-altitude business platform and the test system through an automated script or an API interface, and the test task data includes image format and text format, specifically:

[0015] Collect users' operation data and usage habits from the low-altitude business platform, and analyze user behavior patterns; establish corresponding business logic based on the business rules and processes of the low-altitude business; and obtain historical test data from the test system.

[0016] Preferably, the method also includes preprocessing the collected test task data, processing missing values, outliers and duplicate data in text format data through data cleaning; labeling the preprocessed test task data, including classifying user behaviors and marking key business logic to obtain labeled test task data; using data augmentation technology to generate new test task data from the labeled test task data, and combining the labeled test task data and the new test task data to fine-tune all parameters of the large language model.

[0017] Preferably, fine-tuning all parameters of the large language model is specifically as follows:

[0018] A hyperparameter combination is randomly selected for preliminary testing to obtain preliminary test results, wherein the hyperparameters include learning rate, batch size, and number of fine-tuning rounds; based on the initial test results, the objective function is fitted using the Bayesian method, wherein the cross entropy function is used as the loss function of the large language model, and the objective function is to minimize the loss function; the next hyperparameter combination is selected for testing based on the pre-set expected improvement to obtain new test results; the new test results are added to the large language model update step and iterated until the maximum number of hyperparameter combination attempts is reached to complete the tuning; after the tuning is completed, the performance of the large language model is evaluated using accuracy, precision, recall, and F1-score as evaluation indicators.

[0019] Preferably, the function screenshots and related texts are analyzed by triggering the test case model through an event-driven mechanism and an API call, specifically:

[0020] An event listener is set in the test system, and the event listener is used to monitor the functional screenshots and related texts uploaded by the tester in real time. When the event listener detects the existence of new uploaded data, it automatically triggers the analysis process of the test case model and generates an API request. The API interface receives the uploaded data and converts it into a format that can be processed by the test case model, and passes the converted uploaded data to the test case model for analysis.

[0021] Preferably, preliminary test cases are automatically generated based on the use case analysis results, specifically preset test case generation rules and templates, and key elements in the analysis results are extracted and automatically filled into corresponding test case fields.

[0022] The present application also provides a test system for a low-altitude business platform based on a new AI model, the system comprising an input module, a use case analysis module, a test case generation module, a test case adjustment module, a script analysis module, an automated script generation module, an automated script adjustment module and an automated test execution module, wherein:

[0023] The input module is used by the tester to upload the functional screenshots and related texts corresponding to the low-altitude business platform;

[0024] The use case analysis module has a built-in test case model, which is used to analyze the function screenshots and related texts to obtain the use case analysis results, wherein the test case model is a large language model after all parameters are fine-tuned, and test task data is collected from the low-altitude business platform and the test system through an automated script or an API interface, and the large language model is fine-tuned with all parameters using the test task data, and the test task data includes user behavior, business logic and historical test data;

[0025] The test case generation module is used to automatically generate preliminary test cases according to the case analysis results;

[0026] The test case adjustment module is used by testers to adjust test cases according to actual needs and automatically save and update;

[0027] The script analysis module has a built-in test script model, which is used to analyze the automatically saved and updated test cases to obtain script analysis results, wherein the test script model is a large language model after all parameters are fine-tuned, and test task data is collected from the low-altitude business platform and the test system through an automated script or an API interface, and the large language model is fine-tuned with all parameters using the test task data, and the test task data includes user behavior, business logic and historical test data;

[0028] The automated script generation module is used to generate an automated test script using a request test framework and Python programming language according to the script analysis result;

[0029] The automated script adjustment module is used by testers to adjust automated test scripts according to actual needs and save updates;

[0030] The automated test execution module is used to run the adjusted automated test script to perform testing, capture and record corresponding test results, generate and store test reports.

[0031] Preferably, the test system is also provided with an event listener and an API interface. The event listener is used to monitor the functional screenshots and related texts uploaded by the tester in real time. When the event listener detects the existence of new uploaded data, it automatically triggers the analysis process and generates an API request. The API interface receives the uploaded data and converts it into a format that can be processed by the test script model, and passes the converted uploaded data to the test script model for analysis.

[0032] The present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, a testing method for a low-altitude business platform based on a new AI model as described in any embodiment of the present application is implemented.

[0033] The present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements a testing method for a low-altitude business platform based on a new AI model as described in any embodiment of the present application.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] 1) The present invention provides a testing method and system for low-altitude business platforms based on a new AI model. By using data augmentation and full parameter fine-tuning of a large language model, the system can automatically generate more accurate test cases and test scripts, and perform automated testing, significantly reducing manual intervention and greatly improving test efficiency.

[0036] 2) The present invention provides a testing method and system for low-altitude business platforms based on a new AI model. By using an event-driven mechanism and API calls, the system can respond to the tester's upload operations in real time, and through the collaborative work of multiple modules, a flexible testing process can be achieved to adapt to different testing needs and scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a method flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0039] The present invention provides the following technical solution: a testing method and system for low-altitude business platforms based on a new AI model.

[0040] Embodiment 1:

[0041] This embodiment provides a testing method for a low-altitude service platform based on a new AI model, the method comprising:

[0042] S1. Collect test task data related to the low-altitude business platform, including user behavior, business logic and historical test data;

[0043] S11, test task data;

[0044] This embodiment collects test task data from the low-altitude business platform and the test system through an automated script or an API interface. The test task data includes image format and text format. Specifically:

[0045] S111. Collecting user operation data and usage habits from the low-altitude business platform and analyzing the user behavior pattern; the operation data includes operation time, operation type (such as login, query, submission, etc.), operation result (success or failure) and other user interactions (such as click, slide, etc.);

[0046] S112. Establish corresponding business logic according to the business rules and processes of low-altitude business; obtain historical test data from the test system; the historical test data includes test case execution records, execution status of each test case (pass, fail, blocked, etc.) and error and exception logs;

[0047] S12, data preprocessing;

[0048] The missing values, abnormal values ​​and duplicate data in the text format of the test task data obtained through data cleaning are removed; the pre-processed test task data is annotated, including the classification of user behavior and the marking of key business logic, where the classification of user behavior includes normal operation, abnormal operation and boundary behavior (such as timeout and input error, etc.), and the key business logic includes main business rules and important decision points, and the annotated test task data is obtained;

[0049] Generate new test task data from the labeled test task data using data augmentation technology. In this embodiment, the data augmentation technology includes but is not limited to random perturbation and data synthesis;

[0050] Combine the annotated test task data with the new test task data to fine-tune all parameters of the large language model;

[0051] S2. Fine-tune all parameters of the large language model using the test task data to obtain a test case model and a test script model, and deploy them in the test system;

[0052] S21, large language model;

[0053] The large language model described in this embodiment selects LLaMA as the basic model. LLaMA is a powerful large language model that has the ability to process complex texts and generate high-quality output. This embodiment selects LLaMA-7B as the specific large language model. It is worth noting that this embodiment does not limit the large language model to LLaMA. Users can make decisions based on actual needs and hardware resources.

[0054] The large language model described in this embodiment starts with a pre-trained LLaMA-7B model and loads the pre-trained weights of the model. LLaMA-7B contains 7 billion parameters and is suitable for processing large-scale text data and complex language tasks.

[0055] The test task data is loaded into the large language model in the Dataset format of TensorFlow or PyTorch. The large language model adopts a full-parameter fine-tuning strategy. Specifically, a set of hyperparameter combinations are randomly selected for preliminary testing. The hyperparameters include learning rate, batch size, and number of fine-tuning rounds. The learning rate is set to 1e-5 or 5e-6 to prevent drastic changes in model parameters. The purpose of the preliminary test is to obtain initial model performance indicators; according to the initial test results, the Bayesian method is used to fit the objective function, wherein the cross entropy function is used as the loss function of the large language model, and the objective function is to minimize the loss function; through Bayesian optimization, the next hyperparameter combination is selected for testing according to the expected improvement to obtain new test results; the new test results are added to the update step of the large language model, and it is iterated until the maximum number of hyperparameter combination attempts is reached and the tuning is completed;

[0056] After the tuning is completed, the performance of the large language model is comprehensively evaluated using accuracy, precision, recall, and F1-score as evaluation indicators;

[0057] In the actual training process, AdamW is selected as the optimizer, and hyperparameters such as learning rate and weight decay are set. Preferably, the hyperparameters set at this time are the tuned hyperparameters; the test task data is input into the LLaMA-7B model for multiple rounds of iterative training; preferably, the iterative training is performed in a multi-card or multi-GPU environment to speed up the training; during the training process, the loss function and model performance are monitored, the loss function is the cross entropy loss, and the model performance includes the accuracy and F1 score, and the hyperparameters are adjusted in time to prevent overfitting or underfitting; in order to improve the convergence effect of the model, the cosine annealing scheduler can be used to dynamically adjust the learning rate;

[0058] S22, test case model;

[0059] The generation and output of the test case model are as follows:

[0060] S221. Test case generation:

[0061] Input the description data of the test task, including user behavior patterns, business logic and other information; the fine-tuned LLaMA-7B model generates corresponding test cases based on the input data. The test cases should cover a variety of test scenarios with high accuracy and coverage; use Beam Search or Sampling generative strategies to generate test cases with moderate length and clear semantics; verify the generated test cases manually or automatically to ensure that they meet the actual test requirements;

[0062] S222. Test case model output:

[0063] The test case model is output in a structured data format, which is JSON or XML and contains information such as the description of the test case, expected results, and execution steps. The data format output by the test case model is seamlessly integrated with the test system to support automated test execution.

[0064] S23, test script model;

[0065] The generation and output of the test script model are as follows:

[0066] S231, Test script generation:

[0067] Input the description data of the test task, including user behavior patterns, business logic and other information; the fine-tuned LLaMA-7B model generates the corresponding test script based on the input data. The test script contains detailed test steps and operation instructions to ensure that the automated test system can be executed smoothly; adopt generative strategies such as Beam Search or Sampling to generate test scripts with moderate length and clear semantics; verify the generated test scripts manually or automatically to ensure that they meet the actual test requirements;

[0068] S232, test script model output:

[0069] The test script model outputs a script file in Python or Bash format, which contains detailed test steps, operation instructions and expected results. The script file output by the test script model is seamlessly integrated with the test system to support automated test execution.

[0070] S24, model deployment;

[0071] Setting up the hardware resources and software environment of the test system, wherein the hardware resources include a high-memory GPU and fast storage to support the operation and reasoning of large language models; and the software environment includes Python, TensorFlow, and PyTorch to ensure that the model can run normally;

[0072] Package the test case model and the test script model and their corresponding related dependencies for easy deployment into the test system; deploy the test case model and the test script model as API services, provide services to the outside world through RESTful API or gRPC interface, and the test system automatically generates test cases and test scripts by calling the API interface. Preferably, during the deployment process, set up a monitoring and logging system to monitor the running status and performance indicators of the test case model and the test script model in real time to ensure the high availability and stability of the test case model and the test script model;

[0073] S3. The tester uploads the functional screenshots and related texts corresponding to the low-altitude business platform in the input module of the test system, analyzes the functional screenshots and related texts through the event-driven mechanism and API call triggering test case model, and obtains the case analysis results;

[0074] Specifically, an event listener is set in the test system, and the event listener is used to monitor the function screenshots and related texts uploaded by the tester in real time. When the event listener detects the existence of new uploaded data, it automatically triggers the analysis process of the test case model and generates an API request. The API interface receives the uploaded data and converts it into a format that can be processed by the test case model, and passes the converted uploaded data to the test case model for analysis;

[0075] S4. Automatically generate preliminary test cases according to the use case analysis results. Specifically, preset test case generation rules and templates, extract key elements from the analysis results and automatically fill them into corresponding test case fields;

[0076] The tester adjusts the test case according to actual needs and automatically saves and updates it. The adjustment includes adding, deleting, and modifying the steps and expected results in the test case.

[0077] S5. Based on the event-driven mechanism and API call, the test case is automatically saved and updated to trigger the test script model to analyze it and obtain the script analysis result. This step is based on the same principle as S3 and will not be repeated here.

[0078] S6. Generate an automated test script using the request test framework and Python programming language based on the script analysis results. The generated automated script is highly reusable and scalable. The tester adjusts the automated test script according to actual needs and saves and updates it.

[0079] S7. Run the adjusted automated test script to perform automated testing, capture and record corresponding test results, generate and store a test report based on the corresponding test results.

[0080] Embodiment 2:

[0081] This embodiment provides a test system for a low-altitude business platform based on a new AI model, the system comprising an input module, a use case analysis module, a test case generation module, a test case adjustment module, a script analysis module, an automated script generation module, an automated script adjustment module, and an automated test execution module, wherein:

[0082] The input module is used by the tester to upload the functional screenshots and related texts corresponding to the low-altitude business platform;

[0083] The use case analysis module has a built-in test case model, which is used to analyze the function screenshots and related texts to obtain the use case analysis results, wherein the test case model is a large language model after all parameters are fine-tuned, and test task data is collected from the low-altitude business platform and the test system through an automated script or an API interface, and the large language model is fine-tuned with all parameters using the test task data, and the test task data includes user behavior, business logic and historical test data;

[0084] The test case generation module is used to automatically generate preliminary test cases according to the case analysis results;

[0085] The test case adjustment module is used by testers to adjust test cases according to actual needs and automatically save and update;

[0086] The script analysis module has a built-in test script model, which is used to analyze the automatically saved and updated test cases to obtain script analysis results, wherein the test script model is a large language model after all parameters are fine-tuned, and test task data is collected from the low-altitude business platform and the test system through an automated script or an API interface, and the large language model is fine-tuned with all parameters using the test task data, and the test task data includes user behavior, business logic and historical test data;

[0087] The automated script generation module is used to generate an automated test script using a request test framework and Python programming language according to the script analysis result;

[0088] The automated script adjustment module is used by testers to adjust automated test scripts according to actual needs and save updates;

[0089] The automated test execution module is used to run the adjusted automated test script to perform testing, capture and record corresponding test results, generate and store test reports.

[0090] Preferably, the test system is also provided with an event listener and an API interface. The event listener is used to monitor the functional screenshots and related texts uploaded by the tester in real time. When the event listener detects the existence of new uploaded data, it automatically triggers the analysis process and generates an API request. The API interface receives the uploaded data and converts it into a format that can be processed by the test script model, and passes the converted uploaded data to the test script model for analysis.

[0091] Embodiment 3:

[0092] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, a testing method for a low-altitude business platform based on a new AI model as described in any one of Embodiment 1 is implemented.

[0093] Embodiment 4:

[0094] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, a testing method for a low-altitude business platform based on a new AI model as described in any one of Embodiment 1 is implemented.

[0095] It is worth noting that the system, device and medium described in the present invention are based on the same inventive concept as the method described in Example 1 of the present invention, and will not be repeated here.

[0096] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

[0097] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, platforms, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0098] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. A testing method for low-altitude business platforms based on a new AI model, characterized in that: The method comprises: Collect test task data related to the low-altitude business platform, including user behavior, business logic and historical test data; use the test task data to fine-tune the large language model, obtain the test case model and test script model, and deploy them in the test system; The tester uploads the functional screenshots and related texts corresponding to the low-altitude business platform in the input module of the test system, analyzes the functional screenshots and related texts through the event-driven mechanism and API call triggering test case model, and obtains the case analysis results; Automatically generate preliminary test cases based on the use case analysis results; testers adjust the test cases based on actual needs and automatically save and update; Based on the event-driven mechanism and API calls, the test cases are automatically saved and updated to trigger the test script model to analyze them and obtain the script analysis results; Generate an automated test script using the request test framework and Python programming language based on the script analysis results, and the tester adjusts the automated test script according to actual needs and saves and updates it; Run the adjusted automated test script to perform automated testing, capture and record the corresponding test results, generate and store a test report based on the corresponding test results.

2. According to claim 1, a testing method for low-altitude business platforms based on a new AI model is characterized in that: The test task data is collected from the low-altitude business platform and the test system through an automated script or API interface. The test task data includes image format and text format. Specifically: Collect users' operation data and usage habits from the low-altitude business platform, analyze and obtain user behavior patterns; establish corresponding business logic according to the business rules and processes of low-altitude business; Get historical test data from the test system.

3. According to claim 2, a testing method for low-altitude business platform based on a new AI model is characterized in that: The method also includes preprocessing the collected test task data, processing missing values, abnormal values ​​and repeated data of text format data through data cleaning; labeling the preprocessed test task data, including classifying user behaviors and marking key business logic to obtain labeled test task data; using data augmentation technology to generate new test task data from the labeled test task data, and combining the labeled test task data and the new test task data to fine-tune all parameters of the large language model.

4. According to claim 3, a testing method for low-altitude business platforms based on a new AI model is characterized in that: The specific steps for fine-tuning all parameters of a large language model are as follows: A hyperparameter combination is randomly selected for preliminary testing to obtain preliminary test results, wherein the hyperparameters include learning rate, batch size, and number of fine-tuning rounds; based on the initial test results, the objective function is fitted using the Bayesian method, wherein the cross entropy function is used as the loss function of the large language model, and the objective function is to minimize the loss function; the next hyperparameter combination is selected for testing based on the pre-set expected improvement to obtain new test results; the new test results are added to the large language model update step and iterated until the maximum number of hyperparameter combination attempts is reached to complete the tuning; after the tuning is completed, the performance of the large language model is evaluated using accuracy, precision, recall, and F1-score as evaluation indicators.

5. According to claim 4, a testing method for low-altitude business platform based on a new AI model is characterized in that: The function screenshots and related texts are analyzed through the event-driven mechanism and API call trigger test case model, specifically: An event listener is set in the test system, and the event listener is used to monitor the functional screenshots and related texts uploaded by the tester in real time. When the event listener detects the existence of new uploaded data, it automatically triggers the analysis process of the test case model and generates an API request. The API interface receives the uploaded data and converts it into a format that can be processed by the test case model, and passes the converted uploaded data to the test case model for analysis.

6. According to claim 5, a testing method for low-altitude business platform based on a new AI model is characterized in that: Automatically generate preliminary test cases based on the use case analysis results, specifically preset test case generation rules and templates, extract key elements from the analysis results and automatically fill them into the corresponding test case fields.

7. A test system for low-altitude business platforms based on a new AI model, characterized in that: The system includes an input module, a use case analysis module, a test case generation module, a test case adjustment module, a script analysis module, an automated script generation module, an automated script adjustment module and an automated test execution module, wherein: The input module is used by the tester to upload the functional screenshots and related texts corresponding to the low-altitude business platform; The use case analysis module has a built-in test case model, which is used to analyze the function screenshots and related texts to obtain the use case analysis results, wherein the test case model is a large language model after all parameters are fine-tuned, and test task data is collected from the low-altitude business platform and the test system through an automated script or an API interface, and the large language model is fine-tuned with all parameters using the test task data, and the test task data includes user behavior, business logic and historical test data; The test case generation module is used to automatically generate preliminary test cases according to the case analysis results; The test case adjustment module is used by testers to adjust test cases according to actual needs and automatically save and update; The script analysis module has a built-in test script model, which is used to analyze the automatically saved and updated test cases to obtain script analysis results, wherein the test script model is a large language model after all parameters are fine-tuned, and test task data is collected from the low-altitude business platform and the test system through an automated script or an API interface, and the large language model is fine-tuned with all parameters using the test task data, and the test task data includes user behavior, business logic and historical test data; The automated script generation module is used to generate an automated test script using a request test framework and Python programming language according to the script analysis result; The automated script adjustment module is used by testers to adjust automated test scripts according to actual needs and save updates; The automated test execution module is used to run the adjusted automated test script to perform testing, capture and record corresponding test results, generate and store test reports.

8. According to claim 7, a test system for low-altitude business platforms based on a new AI model is characterized in that: The test system is also provided with an event listener and an API interface. The event listener is used to monitor the functional screenshots and related texts uploaded by the tester in real time. When the event listener detects the existence of new uploaded data, it automatically triggers the analysis process and generates an API request. The API interface receives the uploaded data and converts it into a format that can be processed by the test script model, and passes the converted uploaded data to the test script model for analysis.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements a testing method for a low-altitude business platform based on a new AI model as described in any one of claims 1 to 6.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, a testing method for a low-altitude business platform based on a new AI model as described in any one of claims 1 to 6 is implemented.

Images