AI-based undisclosed vulnerability test early warning method and system, intelligent terminal and storage medium

Abstract

The invention relates to the technical field of vulnerability testing, in particular to an undisclosed vulnerability test early warning method and system based on AI, an intelligent terminal and a storage medium, and the method comprises the following steps: S1, performing to-be-tested function division on to-be-tested software, and matching test instruction sets of different to-be-tested functions in a preset test database; s2, sorting based on the detection result weight values of the test instruction sets to determine test instruction set sorting with vulnerability test reliability from high to low; s3, indicating a preset AI test network model to test a to-be-tested function of the software based on the test instruction set sequence, and marking the instruction set with vulnerabilities in a test result to determine a vulnerability execution instruction; and S4, generating a vulnerability log according to the vulnerability execution instruction and sending out an optimization early warning prompt. The method and the device have the effect of improving the detection of the position vulnerability of the software position so as to facilitate the software optimization early warning.

Description

Technical Field

[0001] The present application relates to the field of vulnerability testing technology, and in particular to an AI-based undisclosed vulnerability testing early warning method, system, intelligent terminal and storage medium. Background Art

[0002] During the software development and testing process, continuous iterative updates are usually required to improve the smoothness and security of basic functions. This process is inseparable from the detection and early warning of vulnerabilities to reduce the impact on normal user use.

[0003] In related technologies, when conducting vulnerability testing and warning on software, it is necessary to collect user usage and feedback as well as continuous testing by software developers, so as to discover vulnerabilities in different functional modules and usage scenarios, and to trace and optimize them.

[0004] Regarding the above-mentioned related technologies, during the iterative optimization process of similar software, when the software operates under different conditions, it will form a variety of scenario impacts, making it impossible for users to know when some unconventional vulnerabilities occur, making it difficult to identify which instructions caused the vulnerability during the tracing process, which is not conducive to early warning and improvement of the vulnerability. Summary of the invention

[0005] In order to improve the testing effect of software location vulnerabilities and facilitate software optimization warning, the present application provides an AI-based undisclosed vulnerability testing and warning method, system, intelligent terminal and storage medium.

[0006] In the first aspect, the present application provides an AI-based undisclosed vulnerability testing and early warning method, which adopts the following technical solutions:

[0007] An AI-based undisclosed vulnerability testing and early warning method, comprising:

[0008] Step S1, dividing the software to be tested into functions to be tested, and matching the test instruction sets of different functions to be tested in a preset test database;

[0009] Step S2, sorting the test instruction sets based on the weight values ​​of the detection results to determine the test instruction sets sorted from high to low in terms of vulnerability test reliability;

[0010] Step S3, based on the test instruction set sorting instruction, the preset AI test network model is instructed to test the software's function to be tested, and the instruction set with vulnerabilities in the test results is marked to determine the vulnerability execution instruction;

[0011] Step S4, generating a vulnerability log according to the vulnerability execution instruction and issuing an optimization warning prompt.

[0012] By adopting the above technical solution, different functions to be tested in the test database are sorted, so that the test instruction sets with higher calibration result accuracy are tested first, so as to quickly find unknown vulnerabilities in the software, and generate corresponding vulnerability logs and early warning prompts, which is convenient for optimization and tracing based on the vulnerability logs, and helps to improve the running stability of the software.

[0013] Optionally, in step S2, when determining the order of the test instruction sets of vulnerability test reliability from high to low, it also includes:

[0014] Search and analyze the weight values ​​of the detection results to determine the adjustable sequence detection instructions with adjacent detection result weight values;

[0015] Perform test time analysis on the reorderable instructions, and swap the order of adjacent reorderable instructions with smaller test time to determine the priority detection instruction set order;

[0016] Updated the priority detection instruction set sorting to test instruction set sorting.

[0017] By adopting the above technical scheme, the weight values ​​of the detection results with similar reliability are compared when sorting the test instruction set, and the detection time of the test instructions in the corresponding test instruction set is analyzed and sorted according to the length of time consumed, so that the test instructions with less time consumption are prioritized, thereby further improving the efficiency of vulnerability detection of the software's function to be tested.

[0018] Optionally, in step S1, when matching the test instruction sets of different functions to be tested in the preset test database, it also includes:

[0019] Analyze the similarity of test instruction sets of different functions to be tested to determine similar functions to be tested whose test instruction set similarity is greater than a preset benchmark similarity;

[0020] Matching shared test files in a preset test database based on similar functions to be tested, and mapping the shared test files and similar functions to be tested to determine the synchronous test functions;

[0021] Test memory is allocated to the synchronous test function according to a preset test memory allocation strategy to determine the number of synchronous tests.

[0022] By adopting the above technical solution, corresponding shared test folders are established for similar functions to be tested with high similarity in test instruction sets, and the test data in the shared test folders are synchronously tested when similar test functions are tested, thereby improving test efficiency and reducing memory usage of test files.

[0023] Optionally, the test memory allocation strategy includes:

[0024] Collect and analyze the remaining instruction sets to be tested and the remaining test memory during the test instruction set test of the same function to be tested;

[0025] Analyze the remaining instruction sets to be tested to determine the test memory required to be deployed when testing different test instructions in the remaining test instruction sets;

[0026] Based on the remaining test memory and the test memory to be allocated, the upper limit number of instruction set tests that can be tested simultaneously is determined;

[0027] Updated the instruction set test upper limit to the number of simultaneous tests.

[0028] By adopting the above technical solution, the test memory is allocated during the testing process of the test instructions so that the software can maintain the maximum upper limit test number when performing the corresponding test, further improving the vulnerability testing efficiency while being less susceptible to the impact of the test memory.

[0029] Optionally, when the instruction set test upper limit is updated to the synchronous test number, the following is also included:

[0030] Calculate the total amount of test memory that needs to be allocated for the instruction set corresponding to the number of simultaneous tests;

[0031] Timing the test end time of the synchronous test instruction set to determine the synchronous test instructions to be tested after the same time ends;

[0032] Perform test memory fluctuation rate analysis based on synchronous test instructions and generate feedback adjustment signals;

[0033] When the test memory fluctuation rate is greater than a preset reference fluctuation rate, the number of synchronous test instructions is reduced according to a feedback adjustment signal.

[0034] By adopting the above technical scheme, the total amount of test memory that needs to be allocated is calculated according to different test instructions, and the test memory fluctuation rate caused by synchronous test instructions is analyzed, so as to generate corresponding feedback adjustment signals to feedback adjust the number of synchronous test instructions, so that synchronous test instructions are not likely to cause insufficient test memory, thereby improving the stability of the test process.

[0035] Optionally, in step S4, issuing the optimization warning prompt also includes:

[0036] Analyze the test instructions according to the vulnerability log to determine whether the function to be tested corresponding to the test instruction is the key function to be tested;

[0037] Match the additional variable test instructions corresponding to the key function to be tested in the preset variable database, and test and verify the key function to be tested with the additional variable test instructions to determine the verification result;

[0038] Based on the verification results, the instruction set with the vulnerability is marked and the vulnerability log is updated.

[0039] By adopting the above technical solution, when an early warning is issued after a vulnerability is generated in a test instruction, the key functions to be tested in the vulnerability log are tested again, and the corresponding attachment variable test instructions are matched, so that the vulnerability test results of the key functions to be tested are more comprehensive.

[0040] Optionally, in step S3, when testing the software function to be tested, the following steps are included:

[0041] Record and store the test parameters in a preset temporary file area, and generate an instruction test process anchor point;

[0042] When the test environment is abnormally closed based on the test process anchor point, the collection software retests instructions;

[0043] Read the test process anchor point according to the retest instruction, and perform the function test to be tested according to the test instruction set;

[0044] When the test instruction set completes the test of the function to be tested, it analyzes whether there is a vulnerability in the function to be tested. If not, the test parameters in the temporary file area without the vulnerability are deleted.

[0045] By adopting the above technical solution, a test process anchor point is generated in the test instructions during the test of the function to be tested, so that when the software is unexpectedly closed during the test process, the test can continue according to the test process anchor point, reducing the time spent on retesting caused by unexpected closure.

[0046] In the second aspect, the present application provides an AI-based undisclosed vulnerability testing and early warning system, which adopts the following technical solutions:

[0047] An AI-based undisclosed vulnerability testing early warning system, including:

[0048] The test instruction matching module divides the software to be tested into functions to be tested and matches the test instruction sets of different functions to be tested in the preset test database;

[0049] A test instruction sorting module is used to sort the test instruction sets based on the weight values ​​of the detection results to determine the order of the test instruction sets from high to low reliability of vulnerability testing;

[0050] The test vulnerability detection module tests the software's functions to be tested based on the test instruction set sorting instructions of the preset AI test network model, and marks the instruction sets with vulnerabilities in the test results to determine the vulnerability execution instructions;

[0051] The vulnerability warning generation module generates vulnerability logs according to vulnerability execution instructions and issues optimization warning prompts.

[0052] By adopting the above technical solution, different functions to be tested in the test database are sorted, so that the test instruction sets with higher calibration result accuracy are tested first, so as to quickly find unknown vulnerabilities in the software, and generate corresponding vulnerability logs and early warning prompts, which is convenient for optimization and tracing based on the vulnerability logs.

[0053] In a third aspect, the present application provides a smart terminal, which adopts the following technical solution:

[0054] A smart terminal includes a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute any of the above-mentioned AI-based undisclosed vulnerability testing and early warning methods.

[0055] By adopting the above technical solution and using smart terminals, the software to be tested can be matched with the corresponding test instruction set, and location vulnerabilities can be detected and eliminated, and vulnerability logs can be generated, which is helpful for software traceability optimization.

[0056] In a fourth aspect, the present application provides a computer storage medium capable of storing corresponding programs, which has the characteristics of improving the accuracy of software testing results for unknown vulnerabilities, and adopts the following technical solutions:

[0057] A computer-readable storage medium stores a computer program that can be loaded by a processor and execute any of the above-mentioned AI-based undisclosed vulnerability testing and early warning methods.

[0058] By adopting the above technical solution, there is a computer program of the undisclosed vulnerability testing and early warning method based on AI in the storage medium. When the computer program is executed, the accuracy of the software's test results for unknown vulnerabilities can be reduced.

[0059] In summary, the present application includes at least one of the following beneficial technical effects:

[0060] 1. Sort the different functions to be tested in the test database so that the test instruction sets with higher test result accuracy are tested first, so as to quickly find out the unknown vulnerabilities in the software and generate corresponding vulnerability logs and early warning prompts, so as to facilitate optimization and tracing according to the vulnerability logs, which helps to improve the running stability of the software;

[0061] 2. Calculate the total amount of test memory that needs to be allocated according to different test instructions, and analyze the test memory fluctuation rate caused by synchronous test instructions, so as to generate corresponding feedback adjustment signals to feedback adjust the number of synchronous test instructions, so that synchronous test instructions are not likely to cause insufficient test memory, so as to improve the stability of the test process;

[0062] 3. When an early warning is issued after a vulnerability is generated in a test instruction, the key functions to be tested in the vulnerability log are tested again, and the corresponding attached variable test instructions are matched, so that the vulnerability test results of the key functions to be tested are more comprehensive. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a method flow chart of steps S1 to S4 in this application.

[0064] Figure 2 It is a method flow chart of steps S201 to S203 in this application.

[0065] Figure 3 It is a method flow chart of steps S101 to S103 in this application.

[0066] Figure 4 It is a method flow chart of steps S1031 to S1034 in this application.

[0067] Figure 5 It is a method flow chart of steps S1035 to S1038 in this application.

[0068] Figure 6 It is a method flow chart of steps S401 to S403 in this application.

[0069] Figure 7 It is a method flow chart of steps S301 to S304 in this application. DETAILED DESCRIPTION

[0070] In order to make the purpose, technical solutions and advantages of this application more clear, the following Figure 1-7 It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0071] The embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings.

[0072] The embodiment of the present application discloses an AI-based undisclosed vulnerability testing and early warning method, which makes the software more efficient in testing and searching for undisclosed vulnerabilities by matching the tested functions of the tested software with corresponding test instruction sets and sorting the test instruction sets according to vulnerability testing reliability. At the same time, a vulnerability log is generated for the found vulnerability test instructions to facilitate subsequent optimization and tracing.

[0073] Reference Figure 1 , a method flow of an undisclosed vulnerability testing early warning method based on AI includes the following steps:

[0074] Step S1: dividing the software to be tested into functions to be tested, and matching the test instruction sets of different functions to be tested in the preset test database;

[0075] The function to be tested is a local function of the software when executing different functions, such as the login module in the music software, the menu module of the control interface, the music recognition module, etc. Different software are set up with corresponding test instruction sets, and the test instruction set stores corresponding test steps, which can check whether the corresponding steps can run normally. Different test instruction sets are stored in a preset test database, and the functions to be tested associated with the corresponding test instruction sets are stored. When the function to be tested is output, the corresponding test instruction set is automatically searched and output.

[0076] Step S2: sorting the test instruction sets based on the weight values ​​of the detection results to determine the test instruction sets sorted from high to low in terms of vulnerability test reliability;

[0077] The detection result weight value indicates the probability of the test instruction set finding a vulnerability in the past test process. For example, the detection result weight value of the first test instruction set finding a vulnerability in the music recognition module is 0.26, indicating that the probability of detecting a vulnerability in the past test process is 26%. At this time, if the weight value is ranked highest in the test instruction set, it means that in the music vulnerability search test, the music recognition module can easily find vulnerabilities. Therefore, by sorting the test instructions in the test instruction set, the efficiency of software testing for unknown vulnerabilities is improved.

[0078] Step S3: Instructing the preset AI test network model to test the software's functions to be tested based on the test instruction set sorting, and marking the instruction set with vulnerabilities in the test results to determine the vulnerability execution instructions;

[0079] The AI ​​test network model is a network data test model pre-established by the staff. It can sort the corresponding test instruction sets to perform relevant tests on the software. By marking the instruction sets that test vulnerabilities, the vulnerabilities of the corresponding functions to be tested can be known. The instructions that test vulnerabilities are defined as vulnerability execution instructions, so that vulnerability optimization can be performed according to the corresponding instructions.

[0080] Step S4: Generate vulnerability logs according to vulnerability execution instructions and issue optimization warning prompts.

[0081] After marking the corresponding vulnerability execution instructions, a vulnerability log is generated, the corresponding test parameters of the vulnerability test instructions are recorded, and an optimization early warning prompt signal is issued to prompt vulnerability optimization.

[0082] Reference Figure 2 In step S2, when determining the order of the test instruction sets of vulnerability test reliability from high to low, it also includes:

[0083] Step S201: searching and analyzing according to the weight values ​​of the detection results to determine the adjustable sequence detection instructions with adjacent detection result weight values;

[0084] Adjacent test instruction sets with similar test result weight values ​​are defined as adjustable sequence test instructions to facilitate their calling in subsequent analysis.

[0085] Step S202: analyzing the test time consumption of the re-sequenceable instructions, and re-ordering the adjacent re-sequenceable instructions with smaller test time consumption to determine the priority detection instruction set order;

[0086] The test time indicates the time required for the test of the adjustable sequence instruction. The test time is determined by pre-establishing a test time database, storing different adjustable sequence instructions in the database, and storing the corresponding test time, so that the corresponding test time is matched and output when the adjustable sequence instruction is output. Based on the queried test time, two adjacent adjustable sequence instructions are sorted, that is, the one with the smaller test time is placed in front, and the one with the larger test time is placed in the back. In this way, the instruction set sorting after adjustment and sorting is determined, and the instruction set sorting is defined as the priority detection instruction set sorting.

[0087] Step S203: Update the priority detection instruction set sequence to the test instruction set sequence.

[0088] By updating the priority detection instruction set to the test instruction set sequence, the vulnerability search can be performed more quickly and effectively when the software to be tested is tested according to the priority detection instruction set sequence.

[0089] Reference Figure 3 In step S1, when matching the test instruction sets of different functions to be tested in the preset test database, it also includes:

[0090] Step S101: analyzing the similarity of test instruction sets of different functions to be tested to determine similar functions to be tested whose test instruction set similarity is greater than a preset reference similarity;

[0091] The test instruction set similarity indicates the similarity of the test instructions. By comparing and analyzing the test instruction sets of the functions to be tested, the test instruction set similarity can be obtained. The benchmark similarity indicates the same similarity. Functions with the same instruction set similarity are marked as the same type of functions to be tested.

[0092] Step S102: matching the shared test files in the preset test database based on the same type of functions to be tested, and mapping the shared test files and the same type of functions to be tested to determine the synchronous test function;

[0093] The shared test file refers to a data file that can be used and tested by different types of functions to be tested. By mapping and associating the shared test file with the same type of functions to be tested, and defining all the associated functions to be tested of the same type as synchronous test functions, it means that the tests can be performed simultaneously.

[0094] Step S103: Allocate test memory for the synchronous test function according to a preset test memory allocation strategy to determine the number of synchronous tests.

[0095] The test memory allocation strategy refers to the strategy for allocating test memory when testing multiple synchronous test functions, so that insufficient memory is less likely to occur during testing, thereby affecting test efficiency or interrupting the test. The number of synchronous tests refers to the number of test instruction sets that can be used for synchronous testing, which is allocated by the test memory allocation strategy.

[0096] Reference Figure 4 , the test memory allocation strategies include:

[0097] Step S1031: collecting and analyzing the remaining instruction sets to be tested and the remaining test memory in the process of testing the instruction sets for executing the same type of functions to be tested;

[0098] During the vulnerability testing of the test instruction set for the same function to be tested, the remaining instruction sets to be tested and the remaining test memory are monitored for easy calling in subsequent further analysis.

[0099] Step S1032: Analyze the remaining instruction sets to be tested to determine the test memory required to be allocated when testing different test instructions in the remaining test instruction sets;

[0100] The test memory to be deployed indicates the test memory required by the remaining instruction sets to be tested, and the test memory to be deployed is obtained by searching for the corresponding test memory in a preset memory instruction database and summing them up.

[0101] Step S1033: Calculate based on the remaining test memory and the test memory to be allocated to determine the upper limit number of instruction set tests that can be performed synchronously;

[0102] By dividing the remaining test memory by the memory to be allocated, the number of test instruction sets that can be tested synchronously can be obtained, and is defined as the upper limit of the instruction set test.

[0103] Step S1034: Update the instruction set test upper limit quantity to the synchronous test quantity.

[0104] By updating the upper limit of the instruction set test quantity to the synchronous test quantity, it is less likely to cause insufficient memory when performing synchronous tests on multiple functions to be tested.

[0105] Reference Figure 5When the instruction set test upper limit is updated to the synchronous test number, it also includes:

[0106] Step S1035: Calculate the total amount of test memory required to be allocated for the instruction set corresponding to the number of synchronous tests;

[0107] The total amount of test memory that needs to be allocated is calculated in order to further analyze and call it during simultaneous testing.

[0108] Step S1036: timing the test end time of the synchronous test instruction set to determine the synchronous test instructions to be tested after the same time ends;

[0109] By monitoring the test end time of the synchronous test instruction set, it is easy to know which instructions are included in the subsequent synchronous test, so as to facilitate further analysis of the allocation stability of the test memory.

[0110] Step S1037: performing test memory fluctuation rate analysis based on the synchronous test instruction and generating a feedback adjustment signal;

[0111] The test memory volatility represents the rate of change of the size of the test memory during allocation, and the feedback adjustment signal represents an adjustment signal generated according to the test memory volatility, so that the allocation of the test memory can be adjusted accordingly, thereby reducing the probability of insufficient memory allocation in the subsequent allocation due to large volatility within the test.

[0112] Step S1038: When the test memory fluctuation rate is greater than a preset reference fluctuation rate, the number of synchronous test instructions is reduced according to the feedback adjustment signal.

[0113] The benchmark volatility is a pre-set volatility value, which means that when allocating test memory, too much test memory will not be lost at the same time. Therefore, by comparing the test memory volatility with the benchmark volatility, if it is greater than, it means that more test memory is used at this time, and a feedback adjustment signal is generated to reduce the synchronous test instructions. Otherwise, a maintenance prompt signal is generated to prompt to maintain the corresponding number of test instructions for testing.

[0114] Reference Figure 6 In step S4, issuing an optimization warning prompt also includes:

[0115] Step S401: analyzing the test instruction according to the vulnerability log to determine whether the function to be tested corresponding to the test instruction is a key function to be tested;

[0116] The key test functions represent the important test functions of the software, which are set by the staff according to the actual situation. By searching and comparing the corresponding test instructions in the vulnerability log, it can be determined whether the function to be tested where the vulnerability occurs is the midpoint test function and marked.

[0117] Step S402: matching the additional variable test instructions corresponding to the key function to be tested in the preset variable database, and testing and verifying the key function to be tested with the additional variable test instructions to determine the verification result;

[0118] When a vulnerability occurs in the key function to be tested, it is necessary to further determine whether there are other vulnerabilities under different environment variables, so as to find the corresponding unknown vulnerabilities more comprehensively.

[0119] Step S403: based on the verification result, mark the instruction set with the vulnerability and update the vulnerability log.

[0120] Reference Figure 7 In step S3, when testing the software function to be tested, it includes:

[0121] Step S301: recording and storing the test parameters in a preset temporary file area, and generating an instruction test process anchor point;

[0122] The test process anchor point indicates the test process point of the test instruction, which can be known by analyzing the test process. The test parameters in the process of testing the function to be tested are temporarily stored in the temporary file area, and the test data can be retained for subsequent calls, and can be deleted later to reduce memory usage.

[0123] Step S302: When the test environment is abnormally closed based on the test process anchor point, the collection software retests instructions;

[0124] When the test environment is shut down due to an abnormality, by collecting the software retest instructions, you can find out whether the software needs to be retested and make corresponding adjustments.

[0125] Step S303: reading the test process anchor point according to the retest instruction, and performing the function test to be tested according to the test instruction set;

[0126] When a retest instruction is collected, the test process of the software can be learned by reading the test process anchor point, and the corresponding test instructions can be executed according to the test process anchor point, so there is no need to retest the test instructions that have been tested, reducing the time spent on vulnerability finding.

[0127] Step S304: when the test instruction set completes the test of the function to be tested, it is analyzed whether there is a vulnerability in the function to be tested, and if not, the test parameters in the temporary file area without the vulnerability are deleted.

[0128] After the function under test is tested, if there is no vulnerability, the test parameters in the temporary file area are deleted to reduce the temporary memory usage of the file.

[0129] Based on the same inventive concept, an embodiment of the present invention provides an AI-based undisclosed vulnerability testing and early warning system, including:

[0130] The test instruction matching module divides the software to be tested into functions to be tested and matches the test instruction sets of different functions to be tested in the preset test database;

[0131] A test instruction sorting module is used to sort the test instruction sets based on the weight values ​​of the detection results to determine the order of the test instruction sets from high to low reliability of vulnerability testing;

[0132] The test vulnerability detection module tests the software's functions to be tested based on the test instruction set sorting instructions of the preset AI test network model, and marks the instruction sets with vulnerabilities in the test results to determine the vulnerability execution instructions;

[0133] The vulnerability warning generation module generates vulnerability logs according to vulnerability execution instructions and issues optimization warning prompts.

[0134] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0135] An embodiment of the present invention provides a computer-readable storage medium storing a computer program that can be loaded by a processor and execute an AI-based undisclosed vulnerability testing and early warning method.

[0136] Computer storage media include, for example, various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0137] Based on the same inventive concept, an embodiment of the present invention provides an intelligent terminal, including a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and execute an AI-based undisclosed vulnerability testing and early warning method.

[0138] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. The specific working process of the system, device and unit described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0139] The above are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Any feature disclosed in this specification (including the abstract and drawings), unless otherwise stated, can be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

Claims

1. An AI-based undisclosed vulnerability testing and early warning method, characterized in that: include: Step S1, dividing the software to be tested into functions to be tested, and matching the test instruction sets of different functions to be tested in a preset test database; Step S2, sorting the test instruction sets based on the weight values ​​of the detection results to determine the test instruction sets sorted from high to low in terms of vulnerability test reliability; Step S3, based on the test instruction set sorting instruction, the preset AI test network model is instructed to test the software's function to be tested, and the instruction set with vulnerabilities in the test results is marked to determine the vulnerability execution instruction; Step S4, generating a vulnerability log according to the vulnerability execution instruction and issuing an optimization warning prompt.

2. According to claim 1, the AI-based undisclosed vulnerability testing and early warning method is characterized in that: In step S2, when determining the order of the test instruction sets of vulnerability test reliability from high to low, it also includes: Search and analyze the weight values ​​of the detection results to determine the adjustable sequence detection instructions with adjacent detection result weight values; Perform test time analysis on the reorderable instructions, and swap the order of adjacent reorderable instructions with smaller test time to determine the priority detection instruction set order; Updated the priority detection instruction set sorting to test instruction set sorting.

3. According to claim 1, the AI-based undisclosed vulnerability testing and early warning method is characterized in that: In step S1, when matching the test instruction sets of different functions to be tested in the preset test database, it also includes: Analyze the similarity of test instruction sets of different functions to be tested to determine similar functions to be tested whose test instruction set similarity is greater than a preset benchmark similarity; Matching shared test files in a preset test database based on similar functions to be tested, and mapping the shared test files and similar functions to be tested to determine the synchronous test functions; Test memory is allocated to the synchronous test function according to a preset test memory allocation strategy to determine the number of synchronous tests.

4. The AI-based undisclosed vulnerability testing and early warning method according to claim 3 is characterized in that: The test memory allocation strategy includes: Collect and analyze the remaining instruction sets to be tested and the remaining test memory during the test instruction set test of the same function to be tested; Analyze the remaining instruction sets to be tested to determine the test memory required to be deployed when testing different test instructions in the remaining test instruction sets; Based on the remaining test memory and the test memory to be allocated, the upper limit number of instruction set tests that can be tested simultaneously is determined; Updated the instruction set test upper limit to the number of simultaneous tests.

5. According to claim 4, the AI-based undisclosed vulnerability testing and early warning method is characterized in that: When the instruction set test upper limit quantity is updated to the simultaneous test quantity, it also includes: Calculate the total amount of test memory that needs to be allocated for the instruction set corresponding to the number of simultaneous tests; Timing the test end time of the synchronous test instruction set to determine the synchronous test instructions to be tested after the same time ends; Perform test memory fluctuation rate analysis based on synchronous test instructions and generate feedback adjustment signals; When the test memory fluctuation rate is greater than a preset reference fluctuation rate, the number of synchronous test instructions is reduced according to a feedback adjustment signal.

6. The AI-based undisclosed vulnerability testing and early warning method according to claim 1, characterized in that: In step S4, issuing an optimization warning prompt also includes: Analyze the test instructions according to the vulnerability log to determine whether the function to be tested corresponding to the test instruction is the key function to be tested; Match the additional variable test instructions corresponding to the key function to be tested in the preset variable database, and test and verify the key function to be tested with the additional variable test instructions to determine the verification result; Based on the verification results, the instruction set with the vulnerability is marked and the vulnerability log is updated.

7. The AI-based undisclosed vulnerability testing and early warning method according to claim 1 is characterized in that: In step S3, when testing the software function to be tested, it includes: Record and store the test parameters in a preset temporary file area, and generate an instruction test process anchor point; When the test environment is abnormally closed based on the test process anchor point, the collection software retests instructions; Read the test process anchor point according to the retest instruction, and perform the function test to be tested according to the test instruction set; When the test instruction set completes the test of the function to be tested, it analyzes whether there is a vulnerability in the function to be tested. If not, the test parameters in the temporary file area without the vulnerability are deleted.

8. An AI-based undisclosed vulnerability testing and early warning system, characterized in that: include: The test instruction matching module divides the software to be tested into functions to be tested and matches the test instruction sets of different functions to be tested in the preset test database; A test instruction sorting module is used to sort the test instruction sets based on the weight values ​​of the detection results to determine the order of the test instruction sets from high to low reliability of vulnerability testing; The test vulnerability detection module tests the software's functions to be tested based on the test instruction set sorting instructions of the preset AI test network model, and marks the instruction sets with vulnerabilities in the test results to determine the vulnerability execution instructions; The vulnerability warning generation module generates vulnerability logs according to vulnerability execution instructions and issues optimization warning prompts.

9. An intelligent terminal, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program that can be loaded by the processor and executes any one of the AI-based undisclosed vulnerability testing and early warning methods as claimed in claims 1 to 7.

10. A computer-readable storage medium, characterized in that: A computer program is stored which can be loaded by a processor and executes any one of the AI-based undisclosed vulnerability testing and early warning methods as claimed in claims 1 to 7.

Images