A software testing method and system based on multiple AI models

By using multiple AI models in software testing, determining test operations and verifying test results, the problems of low testing efficiency and poor results in the existing technology are solved, and more efficient and authentic software testing is achieved.

CN118642972BActive Publication Date: 2024-11-12SHENZHEN SHANGMI NETWORK TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411124406.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-12
Estimated Expiration
2044-08-16

AI Technical Summary

Technical Problem

The prior art fails to effectively utilize multiple AI models in software testing to assist and monitor multiple stages of the overall software testing process, resulting in low testing efficiency and poor results.

Method used

Using a software testing method based on multiple AI models, by obtaining the current testing stage and software information, determining the test operation based on the first AI model, and verifying it based on the second AI model through screen recording and screen recording information of other stages. Improve the degree of automation and efficiency of testing.

Benefits of technology

It improves the automation and efficiency of software testing, enhances the authenticity of test results, and reduces the occurrence of subsequent software errors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118642972B_ABST
    Figure CN118642972B_ABST
Patent Text Reader

Abstract

The present invention discloses a software testing method and system based on multiple AI models, the method comprising: in response to a test start request, obtaining software information of the current test phase and the target test software; based on the current test phase and the software information, determining the test operation corresponding to the current test phase according to the first AI model; executing the test operation on the target test software, and obtaining the screen recording information and test results corresponding to the execution process; based on the screen recording information and the screen recording information corresponding to at least one other test phase, verifying the test results corresponding to the current test phase based on the second AI model. It can be seen that the present invention can improve the automation and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a software testing method and system based on multiple AI models. Background Art

[0002] In the current software testing field, automated testing has become the mainstream of system testing development. Now many computer companies at home and abroad have developed their own automated testing software for automated testing of products, and have also formulated standardized automated testing processes for products. Existing technologies have begun to introduce neural network algorithm technology to assist in the implementation of automated testing, but most of them can only assist in a single test phase or test target, and do not consider using multiple AI models to assist and monitor the entire process at multiple stages in the overall software testing process. Therefore, its testing efficiency is low and the testing effect is poor. It can be seen that the existing technology has defects that need to be solved urgently. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a software testing method and system based on multiple AI models, which can improve the automation and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0004] In order to solve the above technical problems, the first aspect of the present invention discloses a software testing method based on multiple AI models, the method comprising:

[0005] In response to the test start request, obtaining software information of the current test phase and the target test software;

[0006] Based on the current test phase and the software information, determining a test operation corresponding to the current test phase according to a first AI model;

[0007] Execute the test operation on the target test software, and obtain screen recording information and test results corresponding to the execution process;

[0008] According to the screen recording record information and the screen recording record information corresponding to at least one other test stage, based on the second AI model, the test result corresponding to the current test stage is verified.

[0009] As an optional implementation, in the first aspect of the present invention, the software information includes one or more of software type, software module composition, software usage, software programming language, software R&D department and software R&D personnel.

[0010] As an optional implementation, in the first aspect of the present invention, determining the test operation corresponding to the current test phase based on the current test phase and the software information according to the first AI model includes:

[0011] Based on the current test phase and the software information, selecting a target neural network from a plurality of preset prediction neural networks;

[0012] The current test phase and the software information are input into the target neural network to obtain operation information corresponding to the current test phase; the target neural network is trained by a training data set including a plurality of training software information and corresponding test phase annotations and test operation information annotations;

[0013] Based on the operation information, a corresponding operation code is selected from a preset operation code library to obtain a test operation.

[0014] As an optional implementation, in the first aspect of the present invention, the step of selecting a target neural network from a plurality of preset prediction neural networks based on the current test phase and the software information comprises:

[0015] Combining the current test phase and the software information into corresponding prompt words based on a preset prompt word project;

[0016] The prompt word is input into the trained LLM knowledge graph model to obtain corresponding standard software testing information; the standard software testing information includes a plurality of preset software testing standard keywords; the LLM knowledge graph model is constructed based on the knowledge graph algorithm through a plurality of preset software testing training materials;

[0017] According to the standard software test information and the correspondence between the preset information and the model library, a target neural model library is determined from a plurality of preset neural network model libraries;

[0018] According to the current test stage and the software information, a target neural network is screened out from a plurality of prediction neural networks in the target neural model library.

[0019] As an optional implementation, in the first aspect of the present invention, the step of selecting a target neural network from a plurality of prediction neural networks in the target neural model library according to the current test phase and the software information comprises:

[0020] For each prediction neural network in the target neural model library, obtaining test software information in a training data set of the prediction neural network;

[0021] Calculating information similarity between the test software information and the software information;

[0022] Obtaining historical test phases of multiple test records in the historical test records of the prediction neural network;

[0023] Calculate the reciprocal of the average value of the stage distance between the historical test stage and the current test stage of all the test records to obtain a stage distance parameter; the stage distance is the distance between the positions of the historical test stage and the current test stage on a preset stage link sequence;

[0024] Calculating the product of the information similarity and the stage distance parameter to obtain the priority parameter of the prediction neural network;

[0025] The prediction neural network with the highest priority parameter is determined as the target neural network.

[0026] As an optional embodiment, in the first aspect of the present invention, the current test phase or the other test phase is a requirement design test, a requirement verification test, a front-end interaction test, a front-end module test, a server connection test, a server function test or an algorithm rule test.

[0027] As an optional implementation, in the first aspect of the present invention, verifying the test result corresponding to the current test stage based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test stage includes:

[0028] Inputting the screen recording information during the execution of the current test phase into a trained image recognition neural network to obtain software output parameters corresponding to the current test phase; the image recognition neural network is trained by a training data set including a plurality of training screen recording information and corresponding software test output parameter annotations;

[0029] Determine the standard output parameter corresponding to the test result according to the test result and the preset correspondence between the test result and the output parameter corresponding to the current test stage;

[0030] Calculating the parameter difference between the standard output parameter and the software output parameter to obtain the difference parameter corresponding to the current test phase;

[0031] The test result corresponding to the current test phase is verified based on the screen recording record information and the difference parameter of the current test phase, and the screen recording record information and the difference parameter corresponding to at least one other test phase.

[0032] As an optional implementation, in the first aspect of the present invention, verifying the test result corresponding to the current test phase according to the screen recording record information and the difference parameter of the current test phase, and the screen recording record information and the difference parameter corresponding to at least one other test phase, includes:

[0033] For at least one other test phase, determining the screen recording record information and the difference parameter corresponding to the other test phase;

[0034] Determine the correlation parameter between the other test phase and the current test phase according to the distance between the positions of the other test phase and the current test phase on the preset phase link sequence, and the preset mathematical correspondence between the distance and the correlation;

[0035] Calculating a first similarity between the difference parameter of the other test phase and the difference parameter of the current test phase;

[0036] Calculating a second similarity between the screen recording information of the other test phase and the screen recording information of the current test phase; the second similarity is the sum of similarities between image frames at multiple corresponding time points of the two screen recording information;

[0037] Calculating the logarithm of the second similarity with the correlation parameter as the base to obtain a correlation representation value;

[0038] Calculate the product of the first similarity and the correlation representation value to obtain the phase parameter corresponding to the other test phase;

[0039] Calculate the summed average of the phase parameters of all the other test phases to obtain a verification reliability parameter corresponding to the test result corresponding to the current test phase; when the verification reliability parameter is lower than a preset parameter threshold, determine that the test result is unreliable.

[0040] A second aspect of an embodiment of the present invention discloses a software testing system based on multiple AI models, the system comprising:

[0041] An acquisition module, used to obtain software information of the current test phase and the target test software in response to a test start request;

[0042] A determination module, configured to determine, based on the current test phase and the software information, a test operation corresponding to the current test phase according to a first AI model;

[0043] An execution module is used to execute the test operation on the target test software and obtain screen recording information and test results corresponding to the execution process;

[0044] A verification module is used to verify the test result corresponding to the current test phase based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test phase.

[0045] As an optional implementation, in the second aspect of the present invention, the software information includes one or more of software type, software module composition, software usage, software programming language, software R&D department and software R&D personnel.

[0046] As an optional implementation, in the second aspect of the present invention, the determination module determines the specific manner of the test operation corresponding to the current test stage based on the current test stage and the software information according to the first AI model, including:

[0047] Based on the current test phase and the software information, selecting a target neural network from a plurality of preset prediction neural networks;

[0048] The current test phase and the software information are input into the target neural network to obtain operation information corresponding to the current test phase; the target neural network is trained by a training data set including a plurality of training software information and corresponding test phase annotations and test operation information annotations;

[0049] Based on the operation information, a corresponding operation code is selected from a preset operation code library to obtain a test operation.

[0050] As an optional implementation, in the second aspect of the present invention, the specific manner in which the determination module selects the target neural network from the preset multiple prediction neural networks based on the current test phase and the software information includes:

[0051] Combining the current test phase and the software information into corresponding prompt words based on a preset prompt word project;

[0052] The prompt word is input into the trained LLM knowledge graph model to obtain corresponding standard software testing information; the standard software testing information includes a plurality of preset software testing standard keywords; the LLM knowledge graph model is constructed based on the knowledge graph algorithm through a plurality of preset software testing training materials;

[0053] According to the standard software test information and the correspondence between the preset information and the model library, a target neural model library is determined from a plurality of preset neural network model libraries;

[0054] According to the current test stage and the software information, a target neural network is screened out from a plurality of prediction neural networks in the target neural model library.

[0055] As an optional implementation, in the second aspect of the present invention, the specific manner in which the determination module selects the target neural network from the multiple prediction neural networks in the target neural model library according to the current test stage and the software information includes:

[0056] For each prediction neural network in the target neural model library, obtaining test software information in a training data set of the prediction neural network;

[0057] Calculating information similarity between the test software information and the software information;

[0058] Obtaining historical test phases of multiple test records in the historical test records of the prediction neural network;

[0059] Calculate the reciprocal of the average value of the stage distance between the historical test stage and the current test stage of all the test records to obtain a stage distance parameter; the stage distance is the distance between the positions of the historical test stage and the current test stage on a preset stage link sequence;

[0060] Calculating the product of the information similarity and the stage distance parameter to obtain the priority parameter of the prediction neural network;

[0061] The prediction neural network with the highest priority parameter is determined as the target neural network.

[0062] As an optional embodiment, in the second aspect of the present invention, the current test phase or the other test phase is a requirement design test, a requirement verification test, a front-end interaction test, a front-end module test, a server connection test, a server function test or an algorithm rule test.

[0063] As an optional implementation, in the second aspect of the present invention, the verification module verifies the specific manner in which the test result corresponding to the current test stage is verified based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test stage, including:

[0064] Inputting the screen recording information during the execution of the current test phase into a trained image recognition neural network to obtain software output parameters corresponding to the current test phase; the image recognition neural network is trained by a training data set including a plurality of training screen recording information and corresponding software test output parameter annotations;

[0065] Determine the standard output parameter corresponding to the test result according to the test result and the preset correspondence between the test result and the output parameter corresponding to the current test stage;

[0066] Calculating the parameter difference between the standard output parameter and the software output parameter to obtain the difference parameter corresponding to the current test phase;

[0067] The test result corresponding to the current test phase is verified based on the screen recording record information and the difference parameter of the current test phase, and the screen recording record information and the difference parameter corresponding to at least one other test phase.

[0068] As an optional implementation, in the second aspect of the present invention, the verification module verifies the specific manner in which the test result corresponding to the current test stage is verified according to the screen recording record information and the difference parameter of the current test stage, and the screen recording record information and the difference parameter corresponding to at least one other test stage, including:

[0069] For at least one other test phase, determining the screen recording record information and the difference parameter corresponding to the other test phase;

[0070] Determine the correlation parameter between the other test phase and the current test phase according to the distance between the positions of the other test phase and the current test phase on the preset phase link sequence, and the preset mathematical correspondence between the distance and the correlation;

[0071] Calculating a first similarity between the difference parameter of the other test phase and the difference parameter of the current test phase;

[0072] Calculating a second similarity between the screen recording information of the other test phase and the screen recording information of the current test phase; the second similarity is the sum of similarities between image frames at multiple corresponding time points of the two screen recording information;

[0073] Calculating the logarithm of the second similarity with the correlation parameter as the base to obtain a correlation representation value;

[0074] Calculate the product of the first similarity and the correlation representation value to obtain the phase parameter corresponding to the other test phase;

[0075] Calculate the summed average of the phase parameters of all the other test phases to obtain a verification reliability parameter corresponding to the test result corresponding to the current test phase; when the verification reliability parameter is lower than a preset parameter threshold, determine that the test result is unreliable.

[0076] The third aspect of the present invention discloses another software testing system based on multiple AI models, the system comprising:

[0077] A memory storing executable program code;

[0078] a processor coupled to the memory;

[0079] The processor calls the executable program code stored in the memory to execute part or all of the steps in the software testing method based on multiple AI models disclosed in the first aspect of the present invention.

[0080] The fourth aspect of the present invention discloses a computer storage medium, which stores computer instructions. When the computer instructions are called, they are used to execute some or all of the steps in the software testing method based on multiple AI models disclosed in the first aspect of the present invention.

[0081] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0082] The present invention can determine the test operation to perform the test based on the current test stage and software information according to the AI ​​model method, and then use the AI ​​model to comprehensively verify the test results according to the screen recording information during the execution process and the screen recording information of other stages, thereby improving the automation level and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0084] Figure 1 It is a flowchart of a software testing method based on multiple AI models disclosed in an embodiment of the present invention.

[0085] Figure 2 It is a structural diagram of a software testing system based on multiple AI models disclosed in an embodiment of the present invention.

[0086] Figure 3 It is a structural schematic diagram of another software testing system based on multiple AI models disclosed in an embodiment of the present invention. DETAILED DESCRIPTION

[0087] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0088] The terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or equipment.

[0089] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0090] The present invention discloses a software testing method and system based on multiple AI models, which can determine the test operation to perform the test based on the current test stage and software information according to the AI ​​model method, and then use the AI ​​model to comprehensively verify the test results according to the screen recording information during the execution process and the screen recording information of other stages, thereby improving the automation and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors. The following are detailed descriptions.

[0091] Embodiment 1

[0092] See also Figure 1 , Figure 1 : is a flowchart of a software testing method based on multiple AI models disclosed in an embodiment of the present invention. Figure 1 The described software testing method based on multiple AI models can be applied to a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). Figure 1 As shown, the software testing method based on multiple AI models may include the following operations:

[0093] 101. In response to a test start request, obtain software information of a current test phase and a target test software.

[0094] 102. Based on the current test phase and software information, determine the test operation corresponding to the current test phase according to the first AI model.

[0095] 103. Execute a test operation on the target test software, and obtain screen recording information and test results corresponding to the execution process.

[0096] 104. Verify the test result corresponding to the current test phase based on the second AI model according to the screen recording information and the screen recording information corresponding to at least one other test phase.

[0097] It can be seen that the above-mentioned embodiments of the invention can determine the test operation to perform the test based on the current test stage and software information according to the AI ​​model method, and then use the AI ​​model to comprehensively verify the test results according to the screen recording information during the execution process and the screen recording information in other stages, thereby improving the automation level and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0098] As an optional embodiment, in the above steps, the software information includes one or more of software type, software module composition, software purpose, software programming language, software R&D department and software R&D personnel.

[0099] It can be seen that through the above optional embodiments, the content of the software information is clarified, and the information of the software itself and the R&D team can be more comprehensively characterized, so as to facilitate the subsequent determination of test operations, assist in improving the degree of automation and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0100] As an optional embodiment, in the above steps, based on the current test phase and the software information, according to the first AI model, determining the test operation corresponding to the current test phase includes:

[0101] Based on the current test stage and software information, a target neural network is selected from multiple preset prediction neural networks;

[0102] Inputting the current test phase and software information into the target neural network to obtain operation information corresponding to the current test phase; optionally, the target neural network is trained by a training data set including a plurality of training software information and corresponding test phase annotations and test operation information annotations;

[0103] Based on the operation information, a corresponding operation code is selected from a preset operation code library to obtain a test operation.

[0104] It can be seen that through the above-mentioned optional embodiments, the most suitable neural network can be screened out from multiple networks to predict operation information, and then the operation code can be selected based on the operation information to facilitate subsequent test execution, thereby improving the automation level and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0105] As an optional embodiment, in the above steps, based on the current test phase and software information, selecting a target neural network from a plurality of preset prediction neural networks includes:

[0106] Based on the preset prompt word project, the current test phase and software information are combined into corresponding prompt words;

[0107] Input the prompt word into the trained LLM knowledge graph model to obtain the corresponding standard software testing information; optionally, the standard software testing information includes a plurality of preset software testing standard keywords; the LLM knowledge graph model is constructed based on the knowledge graph algorithm through a plurality of preset software testing training materials;

[0108] According to the standard software test information and the correspondence between the preset information and the model library, a target neural model library is determined from among the preset multiple neural network model libraries;

[0109] According to the current test stage and software information, a target neural network is screened out from multiple prediction neural networks in a target neural model library.

[0110] It can be seen that through the above optional embodiments, after the prompt words are determined through the prompt word engineering, the software testing requirements can be standardized based on the trained LLM knowledge graph model to screen out the corresponding model library, and subsequently a more suitable prediction neural network can be accurately determined to facilitate the precise determination of subsequent test operations, assist in improving the automation level and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0111] As an optional embodiment, in the above steps, selecting a target neural network from a plurality of prediction neural networks in a target neural model library according to the current test phase and software information includes:

[0112] For each prediction neural network in the target neural model library, obtaining the test software information in the training data set of the prediction neural network;

[0113] Calculate the information similarity between the test software information and the software information;

[0114] Obtaining historical test phases of multiple test records in the historical test records of the prediction neural network;

[0115] Calculate the reciprocal of the average value of the phase distance between the historical test phase and the current test phase of all test records to obtain a phase distance parameter; optionally, the phase distance is the distance between the positions of the historical test phase and the current test phase on a preset phase link sequence;

[0116] Calculate the product of information similarity and stage distance parameter to obtain the priority parameters of the prediction neural network;

[0117] The prediction neural network with the highest priority parameter is determined as the target neural network.

[0118] It can be seen that through the above-mentioned optional embodiments, the degree of adaptation of each predictive neural network can be determined by calculating the similarity or distance between the historical test software information of the predictive neural network and the historical test stage and the current information, so as to screen out the optimal target neural network, so as to facilitate the precise determination of subsequent test operations, assist in improving the automation level and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0119] As an optional embodiment, in the above steps, the current test phase or other test phases are requirement design test, requirement verification test, front-end interaction test, front-end module test, server connection test, server function test or algorithm rule test.

[0120] It can be seen that through the above optional embodiments, the type of test phase is clarified, which can more comprehensively characterize the characteristics and features of software testing, facilitate subsequent verification of test results, assist in improving the degree of automation and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0121] As an optional embodiment, in the above steps, verifying the test result corresponding to the current test phase based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test phase includes:

[0122] Input the screen recording information during the execution of the current test phase into the trained image recognition neural network to obtain the software output parameters corresponding to the current test phase; optionally, the image recognition neural network is trained by a training data set including a plurality of training screen recording information and corresponding software test output parameter annotations;

[0123] According to the test results and the correspondence between the test results and the output parameters corresponding to the preset current test stage, determine the standard output parameters corresponding to the test results;

[0124] Calculate the parameter difference between the standard output parameter and the software output parameter to obtain the difference parameter corresponding to the current test phase;

[0125] The test result corresponding to the current test phase is verified based on the screen recording record information and difference parameters of the current test phase, and the screen recording record information and difference parameters corresponding to at least one other test phase.

[0126] It can be seen that through the above-mentioned optional embodiments, the output parameters of the screen recording of the test execution can be identified through the image recognition neural network to calculate the difference between the output parameters and the standard parameters, so as to characterize the execution deviation of the automated test, and then verify the results based on the information of multiple other stages, thereby improving the automation level and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0127] As an optional embodiment, in the above steps, verifying the test result corresponding to the current test phase according to the screen recording record information and difference parameters of the current test phase, and the screen recording record information and difference parameters corresponding to at least one other test phase, includes:

[0128] For at least one other test phase, determining the screen recording record information and difference parameters corresponding to the other test phase;

[0129] Determine the correlation parameter between the other test phase and the current test phase according to the distance between the positions of the other test phase and the current test phase on the preset phase link sequence and the preset mathematical correspondence between the distance and the correlation;

[0130] Calculating a first similarity between the difference parameter of the other test phase and the difference parameter of the current test phase;

[0131] Calculate the second similarity between the screen recording record information of the other test phase and the screen recording record information of the current test phase; optionally, the second similarity is the sum of similarities between image frames at multiple corresponding time points of the two screen recording record information;

[0132] Calculate the logarithm of the second similarity with the correlation parameter as the base to obtain a correlation representation value;

[0133] Calculate the product of the first similarity and the correlation representation value to obtain the phase parameter corresponding to the other test phase;

[0134] The summed average of the phase parameters of all other test phases is calculated to obtain the verification reliability parameter corresponding to the test result corresponding to the current test phase; when the verification reliability parameter is lower than a preset parameter threshold, the test result is determined to be unreliable.

[0135] It can be seen that through the above-mentioned optional embodiments, based on the correlation between other stages and the current stage, by comparing the degree of deviation of automated tests between different stages and the degree of similarity of screen recording operations, it is possible to determine whether the screen recording operation similarity between stages with higher correlation in other stages and the current stage is higher, and whether the degree of deviation of automated test execution in all stages remains consistent, so as to verify whether there is human intervention or automated cheating in the test results of the current stage, thereby improving the degree of automation and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0136] Embodiment 2

[0137] See also Figure 2 , Figure 2 : is a schematic diagram of a software testing system based on multiple AI models disclosed in an embodiment of the present invention. Figure 2 The described software testing system based on multiple AI models can be applied to a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). Figure 2 As shown, the software testing system based on multiple AI models may include:

[0138] The acquisition module 201 is used to acquire the software information of the current test phase and the target test software in response to the test start request.

[0139] The determination module 202 is used to determine the test operation corresponding to the current test phase based on the current test phase and software information according to the first AI model.

[0140] The execution module 203 is used to execute the test operation on the target test software and obtain the screen recording information and test results corresponding to the execution process.

[0141] The verification module 204 is used to verify the test results corresponding to the current test phase based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test phase.

[0142] It can be seen that the above-mentioned embodiments of the invention can determine the test operation to perform the test based on the current test stage and software information according to the AI ​​model method, and then use the AI ​​model to comprehensively verify the test results according to the screen recording information during the execution process and the screen recording information in other stages, thereby improving the automation level and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0143] As an optional embodiment, the software information includes one or more of software type, software module composition, software purpose, software programming language, software R&D department and software R&D personnel.

[0144] It can be seen that through the above optional embodiments, the content of the software information is clarified, and the information of the software itself and the R&D team can be more comprehensively characterized, so as to facilitate the subsequent determination of test operations, assist in improving the degree of automation and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0145] As an optional embodiment, the determination module determines the specific manner of the test operation corresponding to the current test phase based on the current test phase and the software information according to the first AI model, including:

[0146] Based on the current test stage and software information, a target neural network is selected from multiple preset prediction neural networks;

[0147] Inputting the current test phase and software information into the target neural network to obtain operation information corresponding to the current test phase; optionally, the target neural network is trained by a training data set including a plurality of training software information and corresponding test phase annotations and test operation information annotations;

[0148] Based on the operation information, a corresponding operation code is selected from a preset operation code library to obtain a test operation.

[0149] It can be seen that through the above-mentioned optional embodiments, the most suitable neural network can be screened out from multiple networks to predict operation information, and then the operation code can be selected based on the operation information to facilitate subsequent test execution, thereby improving the automation level and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0150] As an optional embodiment, the specific manner in which the determination module selects the target neural network from a plurality of preset prediction neural networks based on the current test phase and software information includes:

[0151] Based on the preset prompt word project, the current test phase and software information are combined into corresponding prompt words;

[0152] Input the prompt word into the trained LLM knowledge graph model to obtain the corresponding standard software testing information; optionally, the standard software testing information includes a plurality of preset software testing standard keywords; the LLM knowledge graph model is constructed based on the knowledge graph algorithm through a plurality of preset software testing training materials;

[0153] According to the standard software test information and the correspondence between the preset information and the model library, a target neural model library is determined from among the preset multiple neural network model libraries;

[0154] According to the current test stage and software information, a target neural network is screened out from multiple prediction neural networks in a target neural model library.

[0155] It can be seen that through the above optional embodiments, after the prompt words are determined through the prompt word engineering, the software testing requirements can be standardized based on the trained LLM knowledge graph model to screen out the corresponding model library, and subsequently a more suitable prediction neural network can be accurately determined to facilitate the precise determination of subsequent test operations, assist in improving the automation level and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0156] As an optional embodiment, the specific manner in which the determination module selects the target neural network from multiple prediction neural networks in the target neural model library according to the current test stage and software information includes:

[0157] For each prediction neural network in the target neural model library, obtaining the test software information in the training data set of the prediction neural network;

[0158] Calculate the information similarity between the test software information and the software information;

[0159] Obtaining historical test phases of multiple test records in the historical test records of the prediction neural network;

[0160] Calculate the reciprocal of the average value of the phase distance between the historical test phase and the current test phase of all test records to obtain a phase distance parameter; optionally, the phase distance is the distance between the positions of the historical test phase and the current test phase on a preset phase link sequence;

[0161] Calculate the product of information similarity and stage distance parameter to obtain the priority parameters of the prediction neural network;

[0162] The prediction neural network with the highest priority parameter is determined as the target neural network.

[0163] It can be seen that through the above-mentioned optional embodiments, the degree of adaptation of each predictive neural network can be determined by calculating the similarity or distance between the historical test software information of the predictive neural network and the historical test stage and the current information, so as to screen out the optimal target neural network, so as to facilitate the precise determination of subsequent test operations, assist in improving the automation level and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0164] As an optional embodiment, the current test phase or other test phases are requirement design test, requirement verification test, front-end interaction test, front-end module test, server connection test, server function test or algorithm rule test.

[0165] It can be seen that through the above optional embodiments, the type of test phase is clarified, which can more comprehensively characterize the characteristics and features of software testing, facilitate subsequent verification of test results, assist in improving the degree of automation and efficiency of software testing, verify the results of software testing to improve the authenticity of the test, and reduce subsequent software errors.

[0166] As an optional embodiment, the verification module verifies the specific manner of the test result corresponding to the current test phase based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test phase, including:

[0167] Input the screen recording information during the execution of the current test phase into the trained image recognition neural network to obtain the software output parameters corresponding to the current test phase; optionally, the image recognition neural network is trained by a training data set including a plurality of training screen recording information and corresponding software test output parameter annotations;

[0168] According to the test results and the correspondence between the test results and the output parameters corresponding to the preset current test stage, determine the standard output parameters corresponding to the test results;

[0169] Calculate the parameter difference between the standard output parameter and the software output parameter to obtain the difference parameter corresponding to the current test phase;

[0170] The test result corresponding to the current test phase is verified based on the screen recording record information and difference parameters of the current test phase, and the screen recording record information and difference parameters corresponding to at least one other test phase.

[0171] It can be seen that through the above-mentioned optional embodiments, the output parameters of the screen recording of the test execution can be identified through the image recognition neural network to calculate the difference between the output parameters and the standard parameters, so as to characterize the execution deviation of the automated test, and then verify the results based on the information of multiple other stages, thereby improving the automation level and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0172] As an optional embodiment, the verification module verifies the specific manner of the test result corresponding to the current test phase according to the screen recording record information and difference parameter of the current test phase, and the screen recording record information and difference parameter corresponding to at least one other test phase, including:

[0173] For at least one other test phase, determining the screen recording record information and difference parameters corresponding to the other test phase;

[0174] Determine the correlation parameter between the other test phase and the current test phase according to the distance between the positions of the other test phase and the current test phase on the preset phase link sequence and the preset mathematical correspondence between the distance and the correlation;

[0175] Calculating a first similarity between the difference parameter of the other test phase and the difference parameter of the current test phase;

[0176] Calculate the second similarity between the screen recording record information of the other test phase and the screen recording record information of the current test phase; optionally, the second similarity is the sum of similarities between image frames at multiple corresponding time points of the two screen recording record information;

[0177] Calculate the logarithm of the second similarity with the correlation parameter as the base to obtain a correlation representation value;

[0178] Calculate the product of the first similarity and the correlation representation value to obtain the phase parameter corresponding to the other test phase;

[0179] The summed average of the phase parameters of all other test phases is calculated to obtain the verification reliability parameter corresponding to the test result corresponding to the current test phase; when the verification reliability parameter is lower than a preset parameter threshold, the test result is determined to be unreliable.

[0180] It can be seen that through the above-mentioned optional embodiments, based on the correlation between other stages and the current stage, by comparing the degree of deviation of automated tests between different stages and the degree of similarity of screen recording operations, it is possible to determine whether the screen recording operation similarity between stages with higher correlation in other stages and the current stage is higher, and whether the degree of deviation of automated test execution in all stages remains consistent, so as to verify whether there is human intervention or automated cheating in the test results of the current stage, thereby improving the degree of automation and efficiency of software testing, verifying the results of software testing to improve the authenticity of the test, and reducing subsequent software errors.

[0181] Embodiment 3

[0182] See also Figure 3 , Figure 3 It is another software testing system based on multiple AI models disclosed in an embodiment of the present invention. Figure 3 The described software testing system based on multiple AI models is applied to a data processing system / data processing device / data processing server (wherein the server includes a local processing server or a cloud processing server). Figure 3As shown, the software testing system based on multiple AI models may include:

[0183] A memory 301 storing executable program codes;

[0184] a processor 302 coupled to the memory 301;

[0185] The processor 302 calls the executable program code stored in the memory 301 to execute the steps of the software testing method based on multiple AI models described in the first embodiment.

[0186] Embodiment 4

[0187] An embodiment of the present invention discloses a computer-readable storage medium storing a computer program for electronic data exchange, wherein the computer program enables a computer to execute the steps of the software testing method based on multiple AI models described in the first embodiment.

[0188] Embodiment 5

[0189] An embodiment of the present invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps of the software testing method based on multiple AI models described in Embodiment 1.

[0190] The above describes specific embodiments of the present specification, and other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily have to be performed in the specific order or sequential order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0191] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0192] For the convenience of description, the above device is described in various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0193] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification may be in the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the embodiments of this specification may be in the form of a computer program product implemented in 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.

[0194] This specification is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0195] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0196] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0197] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0198] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0199] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0200] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0201] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0202] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0203] Finally, it should be noted that the software testing method and system based on multiple AI models disclosed in the embodiments of the present invention only disclose the preferred embodiments of the present invention, which are only used to illustrate the technical solution of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments can still be modified, or some of the technical features can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A software testing method based on multiple AI models, characterized in that: The method comprises: In response to the test start request, obtaining software information of the current test phase and the target test software; Based on the current test phase and the software information, determining a test operation corresponding to the current test phase according to a first AI model; Execute the test operation on the target test software, and obtain screen recording information and test results corresponding to the execution process; Verifying the test result corresponding to the current test phase based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test phase, including: Inputting the screen recording information during the execution of the current test phase into a trained image recognition neural network to obtain software output parameters corresponding to the current test phase; the image recognition neural network is trained by a training data set including a plurality of training screen recording information and corresponding software test output parameter annotations; Determine the standard output parameter corresponding to the test result according to the test result and the preset correspondence between the test result and the output parameter corresponding to the current test stage; Calculating the parameter difference between the standard output parameter and the software output parameter to obtain the difference parameter corresponding to the current test phase; For at least one other test phase, determining the screen recording record information and the difference parameter corresponding to the other test phase; Determine the correlation parameter between the other test phase and the current test phase according to the distance between the positions of the other test phase and the current test phase on the preset phase link sequence, and the preset mathematical correspondence between the distance and the correlation; Calculating a first similarity between the difference parameter of the other test phase and the difference parameter of the current test phase; Calculating a second similarity between the screen recording information of the other test phase and the screen recording information of the current test phase; the second similarity is the sum of similarities between image frames at multiple corresponding time points of the two screen recording information; Calculating the logarithm of the second similarity with the correlation parameter as the base to obtain a correlation representation value; Calculate the product of the first similarity and the correlation representation value to obtain the phase parameter corresponding to the other test phase; Calculate the summed average of the phase parameters of all the other test phases to obtain a verification reliability parameter corresponding to the test result corresponding to the current test phase; when the verification reliability parameter is lower than a preset parameter threshold, determine that the test result is unreliable.

2. The software testing method based on multiple AI models according to claim 1, characterized in that: The software information includes one or more of software type, software module composition, software purpose, software programming language, software R&D department and software R&D personnel.

3. The software testing method based on multiple AI models according to claim 2, characterized in that: The determining, based on the current test phase and the software information, according to the first AI model, a test operation corresponding to the current test phase includes: Based on the current test phase and the software information, selecting a target neural network from a plurality of preset prediction neural networks; The current test phase and the software information are input into the target neural network to obtain operation information corresponding to the current test phase; the target neural network is trained by a training data set including a plurality of training software information and corresponding test phase annotations and test operation information annotations; Based on the operation information, a corresponding operation code is selected from a preset operation code library to obtain a test operation.

4. The software testing method based on multiple AI models according to claim 3 is characterized in that: The step of selecting a target neural network from a plurality of preset prediction neural networks based on the current test phase and the software information includes: Combining the current test phase and the software information into corresponding prompt words based on a preset prompt word project; The prompt word is input into the trained LLM knowledge graph model to obtain corresponding standard software testing information; the standard software testing information includes a plurality of preset software testing standard keywords; the LLM knowledge graph model is constructed based on the knowledge graph algorithm through a plurality of preset software testing training materials; According to the standard software test information and the correspondence between the preset information and the model library, a target neural model library is determined from a plurality of preset neural network model libraries; According to the current test stage and the software information, a target neural network is screened out from a plurality of prediction neural networks in the target neural model library.

5. The software testing method based on multiple AI models according to claim 4 is characterized in that: The step of selecting a target neural network from a plurality of prediction neural networks in the target neural model library according to the current test phase and the software information includes: For each prediction neural network in the target neural model library, obtaining test software information in a training data set of the prediction neural network; Calculating information similarity between the test software information and the software information; Obtaining historical test phases of multiple test records in the historical test records of the prediction neural network; Calculate the reciprocal of the average value of the stage distance between the historical test stage and the current test stage of all the test records to obtain a stage distance parameter; the stage distance is the distance between the positions of the historical test stage and the current test stage on a preset stage link sequence; Calculating the product of the information similarity and the stage distance parameter to obtain the priority parameter of the prediction neural network; The prediction neural network with the highest priority parameter is determined as the target neural network.

6. The software testing method based on multiple AI models according to claim 1, characterized in that: The current test phase or the other test phases are requirement design test, requirement verification test, front-end interaction test, front-end module test, server connection test, server function test or algorithm rule test.

7. A software testing system based on multiple AI models, characterized in that: The system comprises: An acquisition module, used to obtain software information of the current test phase and the target test software in response to a test start request; A determination module, configured to determine, based on the current test phase and the software information, a test operation corresponding to the current test phase according to a first AI model; An execution module is used to execute the test operation on the target test software and obtain screen recording information and test results corresponding to the execution process; A verification module, used to verify the test result corresponding to the current test phase based on the second AI model according to the screen recording record information and the screen recording record information corresponding to at least one other test phase, including: Inputting the screen recording information during the execution of the current test phase into a trained image recognition neural network to obtain software output parameters corresponding to the current test phase; the image recognition neural network is trained by a training data set including a plurality of training screen recording information and corresponding software test output parameter annotations; Determine the standard output parameter corresponding to the test result according to the test result and the preset correspondence between the test result and the output parameter corresponding to the current test stage; Calculating the parameter difference between the standard output parameter and the software output parameter to obtain the difference parameter corresponding to the current test phase; For at least one other test phase, determining the screen recording record information and the difference parameter corresponding to the other test phase; Determine the correlation parameter between the other test phase and the current test phase according to the distance between the positions of the other test phase and the current test phase on the preset phase link sequence, and the preset mathematical correspondence between the distance and the correlation; Calculating a first similarity between the difference parameter of the other test phase and the difference parameter of the current test phase; Calculating a second similarity between the screen recording information of the other test phase and the screen recording information of the current test phase; the second similarity is the sum of similarities between image frames at multiple corresponding time points of the two screen recording information; Calculating the logarithm of the second similarity with the correlation parameter as the base to obtain a correlation representation value; Calculate the product of the first similarity and the correlation representation value to obtain the phase parameter corresponding to the other test phase; Calculate the summed average of the phase parameters of all the other test phases to obtain a verification reliability parameter corresponding to the test result corresponding to the current test phase; when the verification reliability parameter is lower than a preset parameter threshold, determine that the test result is unreliable.

8. A software testing system based on multiple AI models, characterized in that: The system comprises: A memory storing executable program code; a processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the software testing method based on multiple AI models as described in any one of claims 1-6.