System detection method and related device

Abstract

The embodiment of the invention discloses a system detection method and a related device. The method comprises the following steps: creating a detection task for a target system; determining a test case used for participating in the detection task and a corresponding standard processing result; determining task configuration information corresponding to the detection task, wherein the task configuration information is used for indicating related execution attributes of the detection task; when the detection task meets the task starting condition, executing the detection task according to the task configuration information; and in the detection task, calling the target system to process the test case to obtain a reference processing result corresponding to the test case, comparing the reference processing result corresponding to the test case with the standard processing result, and determining a detection result corresponding to the target system according to the comparison result. According to the method, whether the system is abnormal or not can be accurately and reliably detected, so that a healthy network environment is better maintained, and the user experience is guaranteed.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular, to a system detection method and related devices. Background Art

[0002] With the rapid development of Internet technology, nowadays, more and more application programs have content publishing functions. When users use the content publishing functions provided by such application programs, they can upload various forms of content such as text, images, and videos for other users of the application program to view.

[0003] In order to avoid the spread of bad content through the above application programs, an audit system is usually set up in the background of the application program. The audit system is used to audit the content uploaded by users and prevent the exposure of bad content when it is found during the audit that the content uploaded by users belongs to bad content. The audit system plays an extremely important role in the operation and maintenance of application programs. Once the audit system malfunctions, a large amount of bad content will be leaked and spread, seriously affecting the network environment and user experience.

[0004] It can be seen that how to accurately and reliably detect whether there is an abnormality in the audit system is an urgent problem to be solved at present, and it is also crucial for maintaining a healthy network environment and ensuring user experience. Summary of the Invention

[0005] Embodiments of this application provide a system detection method and related devices, which can accurately and reliably detect whether there is an abnormality in the audit system, so as to better maintain a healthy network environment and ensure user experience.

[0006] The first aspect of this application provides a system detection method, and the method includes:

[0007] Create a detection task for the target system;

[0008] Determine the test cases to participate in the detection task, and determine the standard processing results corresponding to the test cases; determine the task configuration information corresponding to the detection task, and the task configuration information is used to indicate the relevant execution attributes of the detection task;

[0009] When the detection task meets the task start condition, execute the detection task according to the task configuration information; in the detection task, call the target system to process the test cases, obtain the reference processing results corresponding to the test cases, compare the reference processing results corresponding to the test cases with the standard processing results, and determine the detection results corresponding to the target system according to the comparison results.

[0010] The second aspect of this application provides a system detection device, and the device includes:

[0011] A task creation module, configured to create a detection task for a target system;

[0012] An information determination module, configured to determine test cases for participating in the detection task, and determine the standard processing results corresponding to the test cases; determine the task configuration information corresponding to the detection task, where the task configuration information is used to indicate relevant execution attributes of the detection task;

[0013] A task execution module, configured to execute the detection task according to the task configuration information when the detection task meets the task start condition; in the detection task, call the target system to process the test cases, obtain the reference processing results corresponding to the test cases, compare the reference processing results corresponding to the test cases with the standard processing results, and determine the detection results corresponding to the target system according to the comparison results.

[0014] A third aspect of the present application provides a computer device, where the device includes a processor and a memory:

[0015] The memory is used to store a computer program;

[0016] The processor is configured to execute the steps of the system detection method as described in the first aspect above according to the computer program.

[0017] A fourth aspect of the present application provides a computer-readable storage medium, where the computer-readable storage medium is used to store a computer program, and the computer program is used to execute the steps of the system detection method as described in the first aspect above.

[0018] A fifth aspect of the present application provides a computer program product or a computer program, where the computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the steps of the system detection method as described in the first aspect above.

[0019] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0020] An embodiment of the present application provides a system detection method. This method proposes a dedicated detection mechanism for the audit system to accurately detect whether there are abnormalities in the audit system, that is, to detect whether the audit system can accurately identify bad content. In this method, first, a detection task for the target system used to perform audit processing is created, and the test cases and the corresponding standard processing results for participating in this detection task are determined, as well as the task configuration information for indicating the relevant execution attributes of this detection task; when this detection task meets the task start condition, the detection task can be executed according to the task configuration information. In this detection task, it is necessary to call the target system to be detected to process the determined test cases to obtain the reference processing results corresponding to the test cases. Then, the reference processing results corresponding to the test cases are compared with the standard processing results, and the detection results corresponding to the target system are determined according to the comparison results. In this way, through the above detection task, the target system to be detected is used to audit the test cases with the standard processing results marked, and the reference processing results are obtained. Furthermore, by comparing whether the reference processing results and the standard processing results are consistent, it is determined whether the current audit ability of the target system is reliable, that is, to judge whether the target system can accurately distinguish normal content and bad content at present. Thus, the target system is used to actually audit the real test cases to achieve the accurate detection of the audit ability of the target system; correspondingly, the target system is operated and maintained according to the detection results obtained by the above method, and a healthy network environment can be maintained more reliably, thereby providing a better Internet experience for users. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the application scenario of the system detection method provided by the embodiment of the present application;

[0022] Figure 2 It is a schematic flowchart of a system detection method provided by the embodiment of the present application;

[0023] Figure 3 It is a schematic flowchart of another system detection method provided by the embodiment of the present application;

[0024] Figure 4 It is a schematic flowchart of yet another system detection method provided by the embodiment of the present application;

[0025] Figure 5 It is a schematic flowchart of the construction process of a use case library provided by the embodiment of the present application;

[0026] Figure 6 It is a schematic diagram of the implementation technical architecture of the system detection method provided by the embodiment of the present application;

[0027] Figure 7 It is a schematic flowchart of the dial test function provided by the embodiment of the present application;

[0028] Figure 8 Another schematic diagram of the construction process of the use case library provided by the embodiment of the present application;

[0029] Figure 9 A schematic diagram of the structure of a system detection device provided by the embodiment of the present application;

[0030] Figure 10 A schematic diagram of the structure of the terminal device provided by the embodiment of the present application;

[0031] Figure 11 A schematic diagram of the structure of the server provided by the embodiment of the present application. Detailed implementation manners

[0032] In order to enable those skilled in the art to better understand the solution of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0033] The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above accompanying drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such used data may be interchanged under appropriate circumstances so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these process, method, product or device.

[0034] In the related art, there is currently no dedicated detection mechanism for the audit system. The anomalies of the audit system are mainly discovered through the following three methods: 1) Discover the anomalies of the audit system based on passive notification information. For example, when receiving a large number of user complaint information, or exposure information from relevant media, or notification and instruction information from relevant regulatory agencies, it can be considered that there are anomalies in the audit system. 2) The internal staff actively conducts inspections to discover the anomalies of the audit system. For example, the internal staff can randomly check the content uploaded by users that has been exposed. If a large number of bad contents are found in the randomly checked content, it can be considered that there are anomalies in the audit system. 3) Set corresponding monitoring indicators for each type of bad content. The monitoring indicators are statistically obtained from the data statistics level according to the occurrence rules of historical bad contents. When it is found that a certain type of bad content audited by the audit system does not reach its corresponding monitoring indicator, it is considered that there are anomalies in the audit system.

[0035] However, none of the above existing detection methods can accurately detect the audit ability of the audit system. For the first method, it relies on the notification information of third-party objects such as users, the media, or regulatory agencies to discover the anomalies of the audit system, and cannot achieve the independent detection of the audit ability of the audit system; moreover, the detection efficiency is low. When it is determined that there are anomalies in the audit system based on the notification information of third-party objects, the bad contents often have spread widely, which has had a greater negative impact on the network environment and user experience. For the second method, the internal staff can often only randomly check a very small part of the content, and the randomly checked results are difficult to accurately reflect whether there are really anomalies in the audit system, and it relies on manual detection, so the detection efficiency is low. For the third method, it relies on the preset monitoring indicators to detect whether there are anomalies in the audit system, and has high requirements for the accuracy of the monitoring indicators. And the monitoring indicators are statistically obtained from the occurrence rules of historical bad contents, and their accuracy is often difficult to guarantee. Therefore, it is often difficult to accurately detect whether there are anomalies in the audit system based on the monitoring indicators.

[0036] In order to accurately detect the audit ability of the audit system, the embodiment of the present application provides a system detection method. In this method, first create a detection task for the target system used to perform audit processing, and determine the test cases and the corresponding standard processing results for participating in the detection task, as well as determine the task configuration information for indicating the relevant execution attributes of the detection task; when the detection task meets the task start condition, the detection task can be executed according to the task configuration information. In this detection task, it is necessary to call the target system to be detected to process the determined test cases to obtain the reference processing results corresponding to the test cases. Then, compare the reference processing results corresponding to the test cases with the standard processing results, and determine the detection results corresponding to the target system according to the comparison results.

[0037] The above method proposes a dedicated automated detection mechanism for the target system to autonomously detect whether there are abnormalities in the target system, that is, to detect whether the target system can accurately identify bad content. Through the created detection tasks, the method uses the target system to be detected to review test cases with pre-labeled standard processing results to obtain reference processing results. Then, by comparing whether the reference processing results are consistent with the standard processing results, it is determined whether the current review ability of the target system is reliable, that is, to judge whether the target system can accurately distinguish normal content from bad content at present. Thus, the target system is used to actually review real test cases to achieve accurate detection of the review ability of the target system. In addition, the method can automatically execute detection tasks without any operation by relevant staff in the detection tasks, which can save relevant resources and improve detection efficiency. Correspondingly, operating and maintaining the target system based on the detection results obtained by the above method can more reliably maintain a healthy network environment, thereby providing a better Internet experience for users.

[0038] It should be noted that the system detection method provided in the embodiments of the present application can be executed by a computer device, which can be a terminal device or a server. Among them, terminal devices include but are not limited to mobile phones, computers, intelligent voice interaction devices, smart home appliances, vehicle-mounted terminals, aircraft, etc. A server can be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud server.

[0039] It should be noted that the information, data, and signals involved in the embodiments of the present application are all authorized by relevant objects or fully authorized by all parties, and the collection, use, and processing of relevant data all comply with relevant laws, regulations, and standards of relevant countries and regions.

[0040] To facilitate understanding of the system detection method provided in the embodiments of the present application, the following takes the execution subject of the system detection method as a server as an example to exemplarily introduce the application scenario of the system detection method. See Figure 1 , Figure 1 is a schematic diagram of the application scenario of the system detection method provided in the embodiments of the present application. As Figure 1 shown, this application scenario includes a terminal device 110 and a server 120, and the terminal device 110 and the server 120 can communicate through a network. Among them, the terminal device 110 faces the target object that initiates a detection task for the target system (hereinafter also referred to as the target review system). Figure 1Taking the terminal device 110 as a computer as an example, in practical applications, the terminal device 110 can also be other forms of terminal devices such as mobile phones and tablet computers, and the embodiments of the present application do not make any limitations in this regard; the server 120 is used to execute the system detection method provided by the embodiments of the present application to execute the detection task initiated by the target object through the terminal device 110, and detect whether there is an abnormality in the target audit system.

[0041] In practical applications, the target object can trigger a detection task initiation operation through the terminal device 110. In response to the detection task initiation operation, the terminal device 110 can generate a detection task creation request and send the detection task creation request to the server 120. Correspondingly, after receiving the detection task creation request, the server 120 can create a detection task.

[0042] Then, the target object can trigger a test case selection operation through the terminal device 110 to select a test case to participate in the above detection task; the terminal device 110 responds to the test case selection operation, generates corresponding test case indication information (used to indicate the test case selected through the test case selection operation), and sends the test case indication information to the server 120. Correspondingly, after receiving the test case indication information, the server 120 can determine the test case participating in the detection task according to the test case indication information, and at the same time determine the standard processing result corresponding to the test case (hereinafter also referred to as the standard audit result).

[0043] In addition, the target object can also trigger a task attribute configuration operation through the terminal device 110 to configure the relevant execution attributes of the above detection task; the terminal device 110 responds to the task attribute configuration operation, generates corresponding task attribute indication information (used to indicate the relevant execution attributes configured for the detection task through the task attribute configuration operation), and sends the task attribute indication information to the server 120. Correspondingly, after receiving the task attribute indication information, the server 120 can determine the task configuration information corresponding to the detection task according to the task attribute indication information.

[0044] When the server 120 determines that the detection task meets the task start condition, the server 120 will automatically execute the detection task according to the above task configuration information. When specifically executing the detection task, the server 120 will call the target audit system to be detected indicated by the task configuration information, audit the selected test case, and obtain the reference processing result corresponding to the test case (hereinafter also referred to as the reference audit result); furthermore, compare the reference audit result corresponding to the selected test case with the standard audit result, and determine the detection result corresponding to the target audit system according to the comparison result, that is, determine whether the target audit system can accurately audit relevant content.

[0045] In this way, according to the task execution attributes set for the target object, the test cases selected by the target object are audited using the target audit system to be detected. By comparing whether the reference audit result determined by the target audit system is consistent with the standard audit result, it is determined whether the current audit ability of the target audit system is reliable, that is, it is judged whether the target audit system can accurately distinguish normal content and bad content at present, so as to accurately detect the audit ability of the target audit system.

[0046] It should be understood that Figure 1 The application scenarios shown are only examples. In actual applications, the system detection method provided by the embodiments of the present application can also be applied to other scenarios, and no limitation is imposed on the application scenarios of the system detection method provided by the embodiments of the present application here.

[0047] The system detection method provided by the present application will be introduced in detail below through method embodiments.

[0048] See Figure 2 , Figure 2 which is a schematic flowchart of the system detection method provided by the embodiments of the present application. For the convenience of description, the following takes the execution subject of this system detection method as a server as an example for introduction. As Figure 2 shown, this system detection method includes the following steps:

[0049] Step 201: Create a detection task for the target system.

[0050] In the embodiments of the present application, the target system can specifically be called the target audit system. When it is necessary to detect whether the audit ability of the target audit system is normal, the server can first respond to relevant requirements and create a detection task for the target audit system.

[0051] It should be noted that the target system in the embodiments of the present application, that is, the target audit system, can be any system used for auditing content. Specifically, it can be used to audit content (such as content published by users) on a specific network platform to identify whether it belongs to inappropriate content that is not suitable for other users to view. The target audit system can specifically be represented as an audit link formed by connecting several intermediate processing nodes in series or in parallel. The above-mentioned intermediate processing nodes are used to execute intermediate processing steps during the audit process, and the final audit result determined by the entire audit link is determined according to the intermediate processing results obtained through each intermediate processing step. In addition, the target audit system in the embodiments of the present application can be used to audit various forms of content. For example, the target audit system can be used to audit text (such as text information published by users, review text information, etc.), images (such as images published by users, review images, user avatars, etc.), videos (such as short videos and long videos published by users, etc.), association relationships (such as the relationship between the content published by a user and the qualification information of the user), etc. The embodiments of the present application do not make any limitations on the audit content of the target audit system here.

[0052] It should be noted that the detection task in the embodiments of the present application is a task used to detect whether the audit ability of the audit system is normal; in this detection task, the audit system to be detected can be called to audit real use cases to obtain corresponding audit results, and then, compare whether the audit results obtained through this audit system are consistent with the standard audit results to determine whether the audit ability of this audit system is normal. The execution method of this detection task will be specifically introduced below, and for details, please refer to the relevant content below.

[0053] As an example, the server can respond to the detection task initiation operation triggered by the target object and create a detection task for the target audit system. Here, the target object is the object that initiates the detection task, and specifically, it can be a staff member with a detection requirement for the target audit system. Specifically, the target object can access a specific website through a terminal device, and this website can provide a service for detecting whether the audit system is abnormal; or, the target object can use a specific application program running on the terminal device, and this application program can provide a service for detecting whether the audit system is abnormal. Then the target object can trigger the detection task initiation operation through the detection task initiation interface on this website or this application program. For example, the target object can click the detection task initiation control on the detection task initiation interface; the terminal device responds to the detection task initiation operation triggered by the target object, can generate a detection task creation request, and send this detection task creation request to the server. After receiving this detection task creation request, the server creates a detection task accordingly.

[0054] Of course, in practical applications, the server can also independently create a detection task for the target audit system; for example, when the preset detection conditions for the target audit system are met, it independently creates a detection task for the target audit system. The detection conditions here can be that the number of complaint messages received for the content on the target network platform exceeds a preset threshold, and the target network platform here is the network platform to which the target audit system is applied. The embodiments of the present application do not make any limitations on the creation method of the detection task.

[0055] Step 202: Determine the test cases to participate in the detection task, and determine the standard processing results corresponding to the test cases; determine the task configuration information corresponding to the detection task, where the task configuration information is used to indicate the relevant execution attributes of the detection task.

[0056] After the server creates a detection task for the target audit system, it is necessary to further determine the test cases to participate in the detection task, that is, to determine the audit objects that the target audit system needs to process in the detection task, and determine the standard processing results corresponding to the test cases. The standard processing results can also be referred to as standard audit results. In addition, the server also needs to determine the task configuration information used to indicate the relevant execution attributes of the detection task.

[0057] It should be noted that the test cases in the embodiments of the present application are the audit objects of the target audit system in the detection task, that is, when the detection task is executed, it is necessary to call the target audit system to audit the test cases. It should be understood that the form of the test cases can be determined according to the form of the content audited by the target audit system. For example, if the target audit system is used to audit text, then the test cases need to be text; if the target audit system is used to audit video, then the test cases need to be video, and so on; specifically, the test cases can be at least one form of content such as text, image, video, association relationship (such as the correspondence between the content published by the user and the qualification information of the user). It should be understood that usually, in order to ensure accurate detection of the audit system, multiple test cases need to be determined.

[0058] The standard processing result corresponding to the test case, that is, the standard review result, is the accurate review result corresponding to the test case. The standard review result corresponding to the test case at least includes the standard final review result, and the standard final review result is used to characterize whether the test case belongs to bad content. More specifically, the standard final review result can also be used to characterize the specific type of bad content to which the test case belongs. Optionally, the standard review result corresponding to the test case may further include at least one standard intermediate processing result, and at least one standard intermediate processing result corresponds to at least one intermediate processing node one by one. For example, when the target review system includes n (n is an integer greater than or equal to 1) intermediate processing nodes, the standard review result may include the standard intermediate processing results corresponding to each of the n intermediate processing nodes, and the standard intermediate processing result is the accurate processing result that should be output by its corresponding intermediate processing node.

[0059] It should be noted that the task configuration information in the embodiments of the present application is information used to indicate the relevant execution attributes of the detection task, and the relevant execution attributes of the detection task are used to indicate the execution manner of the detection task. Exemplarily, the task configuration information may specifically include at least one of call interface information, task execution frequency information, task scenario information, detection concurrency number, and task warning information.

[0060] Among them, the call interface information is used to indicate the interface used to call the target audit system when performing the detection task. The task execution frequency information is used to indicate whether the executed detection task is a one-time task or a periodic task; when the task execution frequency information indicates that the detection task is a one-time task, the task configuration information may further include start time information for indicating the start time of the detection task; when the task execution frequency information indicates that the detection task is a periodic task, the task configuration information may further include start time information for indicating the start time of the first detection task, and task cycle information for indicating the execution cycle of the detection task. The task scenario information is used to indicate the audit scenario to which the target audit system to be detected applies, such as being applied to detect the content published by users, being applied to detect the content of user comments, being applied to detect user avatars, etc. In different audit scenarios, the audit criteria of the target audit system may be different. The detection concurrency is used to indicate the number of test cases audited by calling the target audit system within a single time unit. For example, the detection concurrency can be used to indicate that the target audit system is called to audit 100 test cases per hour; by setting the detection concurrency, it is possible to prevent the target audit system from consuming too many resources to audit test cases and affecting the normal business of the target audit system. The task alarm information can be used to indicate the alarm method, alarm object, and alarm content template. When it is detected that there is an abnormality in the target audit system, the alarm content generated based on the alarm content template can be sent to the alarm object indicated by the task alarm information (such as the relevant responsible person of the target audit system) through the alarm method indicated by the task alarm information (such as phone call, text message, email, instant messaging application, etc.).

[0061] As an example, the server can determine the test cases participating in the detection task in response to the test case selection operation triggered by the target object. Specifically, after the target object triggers the detection task initiation operation through the detection task initiation interface of the website or application, the website or application can jump to display the test case selection interface. Furthermore, the target object can trigger the test case selection operation through the test case selection interface to select the test cases used to participate in the detection task in the pre-built use case library; the terminal device responds to the test case selection operation, generates the corresponding test case indication information (which is used to indicate the test cases selected through the test case selection operation), and sends the test case indication information to the server; after receiving the test case indication information, the server can determine the test cases selected by the target object in the use case library according to the test case indication information, and retrieve the corresponding standard audit results of the selected test cases from the use case library.

[0062] Of course, in practical applications, the server can also independently determine the test cases to participate in the detection task. For example, the server can randomly select test cases of this form from the use case library as the test cases to participate in the detection task according to the form of the content audited by the target audit system, and retrieve the corresponding standard audit results of the selected test cases from the use case library. The embodiments of the present application do not make any limitations on the method of determining the test cases to participate in the detection task here.

[0063] As an example, the server can determine the task configuration information corresponding to the detection task in response to the task attribute configuration operation triggered by the target object. Specifically, after the target object completes the test case selection operation through the test case selection interface, the website or application can further jump to display the task attribute configuration interface. The target object can trigger the task attribute configuration operation through this task attribute configuration interface to set the relevant execution attributes of the detection task, such as setting the call interface, execution frequency, task scenario, detection concurrency, alarm method, alarm object, alarm content template, etc. of the detection task; the terminal device responds to this task attribute configuration operation, generates a corresponding task attribute indication information (which is used to indicate the execution attributes configured through the task attribute configuration operation), and sends this task attribute indication information to the server; after receiving this task attribute indication information, the server can accordingly determine the task configuration information corresponding to the detection task according to this task attribute indication information.

[0064] It should be understood that in practical applications, the target object can also first configure the execution attributes of the detection task and then select the test cases to participate in the detection task. The embodiments of the present application do not make any limitations on the order of determining the task configuration information and the test cases here.

[0065] Step 203: When the detection task meets the task start condition, execute the detection task according to the task configuration information; in the detection task, call the target system to process the test case to obtain the reference processing result corresponding to the test case, compare the reference processing result corresponding to the test case with the standard processing result, and determine the detection result corresponding to the target system according to the comparison result.

[0066] When the above detection task meets the task start condition, the server can execute the detection task according to the pre-determined task configuration information. It should be understood that the task start condition here is the condition that triggers the start of the execution of the detection task; the task start condition can be detecting that the target object triggers a task start operation for the detection task. For example, detecting that the target object triggers a click operation on the task start control is considered that the detection task meets the task start condition; the task start condition can also be determined according to the task configuration information. For example, when the task configuration information includes start time information indicating the start time of the detection task, it can be determined that the current time reaches the start time indicated by the start time information as the condition for meeting the task start condition. Or, when the task configuration information includes task cycle information indicating the execution cycle of the detection task, it can be determined that the time interval between the current time and the execution time of the previous detection task reaches the execution cycle indicated by the task cycle information as the condition for meeting the task start condition.

[0067] When specifically executing the detection task, the server can, according to the task configuration information, correspondingly call the target audit system to be detected to audit the pre-determined test case, and obtain the reference processing result corresponding to the test case. The reference processing result can also be called the reference audit result; it should be understood that the reference processing result corresponding to the test case, that is, the reference audit result, is the audit result determined by the target audit system to be audited for the test case. The reference audit result may be accurate or inaccurate; the reference audit result includes the reference final audit result, that is, the final audit result determined by the target audit system, which is used to characterize whether the test case belongs to bad content, and more specifically, is used to characterize the type of bad content to which the test case belongs; optionally, the reference audit result may further include at least one reference intermediate processing result. The at least one reference intermediate processing result corresponds one-to-one with at least one intermediate processing node in the target audit system. The reference intermediate processing result is the processing result determined by its corresponding intermediate processing node for the test case.

[0068] After the server obtains the reference audit result corresponding to the test case, it can compare the reference audit result corresponding to the test case with the standard audit result corresponding to the test case to obtain the comparison result corresponding to the test case. Furthermore, the server can determine the detection result corresponding to the target audit system according to the comparison results corresponding to each test case used in the detection task, that is, determine whether the audit ability of the target audit system is normal and whether the target audit system can accurately audit bad content.

[0069] The above method proposes a dedicated automated detection mechanism for the target review system to autonomously detect whether there are abnormalities in the target review system, that is, to detect whether the target review system can accurately identify bad content. Through the created detection tasks, this method uses the target review system to be detected to review test cases with pre-labeled standard review results, obtaining reference review results. Then, by comparing whether the reference review results are consistent with the standard review results, it determines whether the current review ability of the target review system is reliable, that is, judges whether the target review system can currently accurately distinguish normal content from bad content. Thus, the target review system is used to actually review real test cases to achieve accurate detection of the review ability of the target review system. In addition, this method can automatically execute detection tasks without any operation by relevant staff in the detection tasks, which can save relevant resources and improve detection efficiency. Correspondingly, operating and maintaining the target review system based on the detection results obtained by the above method can more reliably maintain a healthy network environment, thereby providing a better Internet experience for users.

[0070] In a possible implementation, the test cases used in the detection tasks can be white-box test cases. A white-box test case refers to a test case for which the processing results of each intermediate processing node in the review system can be obtained. Correspondingly, the standard review results corresponding to the test case can include at least one standard intermediate processing result and a standard final processing result; the at least one standard intermediate processing result corresponds one-to-one with at least one intermediate processing node in the review system, and the standard intermediate processing result is the accurate processing result determined by its corresponding intermediate processing node for the test case; the standard final processing result is the accurate review result finally determined for the test case, used to represent whether the test case belongs to bad content and even used to represent the type of bad content to which the test case belongs.

[0071] In this case, refer to Figure 3 , Figure 3 which is a schematic flowchart of a system detection method provided by an embodiment of this application. As Figure 3 shown, the "call the target system to process the test case and obtain the reference processing result corresponding to the test case" in step 203 above includes:

[0072] Step 301: Call the target review system to review the test case, obtaining the reference intermediate processing result determined by the intermediate processing node in the target review system for the test case and the reference final review result finally output by the target review system for the test case.

[0073] For test cases belonging to white-box test cases, when the server calls the target audit system to audit them, it will obtain the reference intermediate processing results determined by the intermediate processing nodes of the target audit system for the test cases. For example, the server can pull at least one reference intermediate processing result determined by at least one intermediate processing node of the target audit system for the test case through an Agent module. It should be understood that the at least one reference intermediate processing result corresponds one-to-one with the at least one intermediate processing node, and the reference intermediate processing result is the processing result output by its corresponding intermediate processing node during the process of auditing the test case.

[0074] In addition, after the target audit system completes the audit operation for the test case, the server can also obtain the reference final audit result finally determined by the target audit system for the test case. For example, the server can pull the audit result finally output by the target audit system for the test case through the proxy module. It should be understood that the reference final audit result is used to characterize whether the test case is bad content, and can even characterize the type of bad content to which the test case belongs.

[0075] Correspondingly, "comparing the reference processing result corresponding to the test case with the standard processing result, and determining the detection result corresponding to the target system according to the comparison result" in step 203 above includes:

[0076] Step 302: Compare whether the reference intermediate processing result and the standard intermediate processing result with a corresponding relationship are consistent to obtain an intermediate comparison result; and compare whether the reference final audit result and the standard final audit result are consistent to obtain a final comparison result.

[0077] Step 303: Determine whether there is an abnormality in the target audit system according to the intermediate comparison result and the final comparison result; in the case of determining that there is an abnormality in the target audit system, determine the abnormal intermediate processing node in the target audit system according to the intermediate comparison result.

[0078] For test cases belonging to white-box test cases, when the server compares the reference audit result corresponding to the test case with the standard audit result, it needs to compare both the reference intermediate processing result corresponding to the test case with the standard intermediate processing result and the reference final audit result corresponding to the test case with the standard final audit result.

[0079] When comparing the reference intermediate processing results corresponding to the comparison test cases with the standard intermediate processing results, the server needs to first determine each pair of intermediate results; each pair of intermediate results includes a reference intermediate processing result and a standard intermediate processing result with a corresponding relationship. The reference intermediate processing result and the standard intermediate processing result with a corresponding relationship mean that the reference intermediate processing result and the standard intermediate processing result correspond to the same intermediate processing node, that is, the reference intermediate processing result and the standard intermediate processing result are obtained through the same intermediate processing steps; it should be understood that the number of the above pairs of intermediate results corresponds to the number of intermediate processing nodes. For each pair of intermediate results, the server will compare whether the reference intermediate processing result in the pair of intermediate results is consistent with the standard intermediate processing result, so as to obtain the intermediate comparison result corresponding to the pair of intermediate results.

[0080] When comparing the reference final review result corresponding to the comparison test case with the standard final review result, the server can directly compare whether the reference final review result is consistent with the standard final review result, so as to obtain the final comparison result. Specifically, if both the reference final review result and the standard final review result only characterize whether the test case belongs to bad content, then, if both the reference final review result and the standard final review result characterize that the test case belongs to bad content, or both characterize that the test case does not belong to bad content, it is considered that the reference final review result and the standard final review result are consistent; otherwise, it is considered that the reference final review result and the standard final review result are inconsistent; if the reference final review result and the standard final review result, in addition to characterizing whether the test case belongs to bad content, also characterize the type of bad content to which the test case belongs, then, in addition to performing the above comparison operation, it is also necessary to compare the types of bad content characterized by the reference final review result and the standard final review result respectively. If the types of bad content to which the test case belongs characterized by the reference final review result and the standard final review result are the same, it is considered that the reference final review result and the standard final review result are consistent; otherwise, it is considered that the reference final review result and the standard final review result are inconsistent.

[0081] Furthermore, the server can determine whether there is an abnormality in the target audit system based on the above intermediate results for their respective corresponding intermediate comparison results and the final comparison result. Exemplarily, for each test case belonging to the white-box use case, if there is at least one comparison result among the intermediate results associated with the test case for their respective corresponding intermediate comparison results and the final comparison result associated with the test case indicating that the compared data is inconsistent, it can be considered that the comparison result associated with the test case indicates an abnormality in the target audit system. On the contrary, if the intermediate results associated with the test case for their respective corresponding intermediate comparison results and the final comparison result associated with the test case all indicate that the compared data is consistent, it can be considered that the comparison result associated with the test case indicates that the target audit system is not abnormal. Furthermore, count the proportion of test cases whose associated comparison results indicate an abnormality in the target audit system among all test cases used in the detection task. If this proportion exceeds the first preset proportion threshold, it can be considered that the target audit system is abnormal. If this proportion does not exceed the first preset proportion threshold, it can be considered that the target audit system is not abnormal.

[0082] In the case of determining that there is an abnormality in the target audit system, the server can also determine the abnormal intermediate processing node of the target audit system, that is, locate the cause of the abnormality in the target audit system, based on the above intermediate results for their respective corresponding intermediate comparison results. Exemplarily, for each intermediate processing node of the target audit system, the server can determine the intermediate result pairs corresponding to the intermediate processing node among the intermediate result pairs associated with each test case, and obtain the intermediate comparison results corresponding to these intermediate result pairs. Then, count the proportion of intermediate comparison results indicating inconsistent compared data among these intermediate comparison results. If this proportion exceeds the second preset proportion threshold, it can be considered that this intermediate processing node is an abnormal intermediate processing node. If this proportion does not exceed the second preset proportion threshold, it can be considered that this intermediate processing node is not an abnormal intermediate processing node.

[0083] In this way, through the above method, use the target audit system to audit the white-box use case, obtain the reference intermediate processing results generated by each intermediate processing node in the audit process and the reference final audit result determined by the target audit system. Then, compare the reference intermediate processing results with the standard intermediate processing results, and compare the reference final audit result with the standard final audit result. Determine whether the target audit system is abnormal based on the obtained comparison results, and can also locate the cause of the abnormality in the target audit system, realizing more refined detection of the target audit system.

[0084] In a possible implementation, the test cases used in the detection task can be black-box test cases, which refer to test cases that can only obtain the final review results. Correspondingly, the standard review result corresponding to the test case is the standard final review result, that is, the accurate review result finally determined for the test case. The standard final review result is used to characterize whether the test case belongs to bad content, and even to characterize the type of bad content to which the test case belongs.

[0085] In this case, refer to Figure 4 , Figure 4 which is a schematic flowchart of a system detection method provided by an embodiment of the present application. As Figure 4 shown, the step of "invoking the target system to process the test case to obtain the reference processing result corresponding to the test case" in step 203 above includes:

[0086] Step 401: Invoke the target review system to review the test case to obtain the reference final review result finally output by the target review system for the test case.

[0087] For a test case that belongs to a black-box test case, when the server invokes the target review system to review it, it will only obtain the reference final review result finally determined by the target review system for the test case. For example, the server can pull the review result finally output by the target review system for the test case through the proxy module. It should be understood that the reference final review result is used to characterize whether the test case belongs to bad content, and even can characterize the type of bad content to which the test case belongs.

[0088] Correspondingly, the step of "comparing the reference processing result corresponding to the test case and the standard processing result, and determining the detection result corresponding to the target review system according to the comparison result" in step 203 above includes:

[0089] Step 402: Compare whether the reference final review result corresponding to the test case is consistent with the standard final review result, and determine the detection result corresponding to the target review system according to the comparison result.

[0090] For a test case that belongs to a black-box test case, when the server compares the reference review result and the standard review result corresponding to the test case, it only needs to compare the reference final review result and the standard final review result corresponding to the test case.

[0091] When comparing the reference final review result corresponding to the comparison test case with the standard final review result, the server can directly compare whether the reference final review result is consistent with the standard final review result to obtain the comparison result corresponding to the test case. Specifically, if both the reference final review result and the standard final review result only represent whether the test case belongs to bad content, then, if both the reference final review result and the standard final review result represent that the test case belongs to bad content, or both represent that the test case does not belong to bad content, it is considered that the reference final review result and the standard final review result are consistent; otherwise, it is considered that the reference final review result and the standard final review result are inconsistent. If the reference final review result and the standard final review result, in addition to representing whether the test case belongs to bad content, also represent the type of bad content to which the test case belongs, then, in addition to performing the above comparison operation, it is also necessary to compare the types of bad content represented by the reference final review result and the standard final review result respectively. If the types of bad content to which the test case belongs represented by the reference final review result and the standard final review result are the same, it is considered that the reference final review result and the standard final review result are consistent; otherwise, it is considered that the reference final review result and the standard final review result are inconsistent.

[0092] Furthermore, the server can determine whether there is an abnormality in the target review system according to the comparison results corresponding to each test case. Exemplarily, the server can count the comparison results corresponding to each test case used in the detection task, determine the proportion of the comparison results indicating inconsistent compared data among all comparison results. If the proportion exceeds a preset ratio threshold, it can be considered that there is an abnormality in the target review system; if the proportion does not exceed the preset ratio threshold, it can be considered that there is no abnormality in the target review system.

[0093] In this way, through the above method, the black-box test cases are reviewed by the target review system to obtain the reference final review result determined by the target review system. Then, the reference final review result is compared with the standard final review result, and whether the target review system is abnormal is determined according to the obtained comparison result, so as to realize the basic detection of the target review system and improve the detection efficiency.

[0094] It should be understood that in practical applications, the test cases used in the detection task can only include white-box test cases, or only include black-box test cases, or can also include both white-box test cases and black-box test cases. For example, the white-box test cases and black-box test cases participating in the detection task are selected according to a specific ratio. The embodiments of the present application do not make any limitations on the types and distributions of the test cases used in the detection task.

[0095] In a possible implementation, when there are multiple test cases participating in the detection task, the step of "comparing the reference processing result corresponding to the test case with the standard processing result and determining the detection result corresponding to the target system according to the comparison result" in step 203 above includes:

[0096] For each of the test cases, compare whether the reference processing result corresponding to the test case is consistent with the standard processing result to obtain the comparison result corresponding to the test case;

[0097] Count the comparison results corresponding to each of the test cases respectively, and determine the proportion of the comparison results indicating inconsistent compared data in all the comparison results;

[0098] When the proportion exceeds a preset proportion threshold, determine that the target system is abnormal.

[0099] Specifically, for each test case used in the detection task, the server can compare whether the reference review result corresponding to the test case is consistent with the standard review result, so as to obtain the comparison result corresponding to the test case. As introduced above, when the test case is a white-box case, when the server compares the reference review result corresponding to the test case with the standard review result, it is necessary to compare the reference intermediate processing result and the standard intermediate processing result with a corresponding relationship, and compare the reference final review result and the standard final review result. If any of the obtained comparison results (including the comparison results of each pair of reference intermediate processing results and standard intermediate processing results, and the comparison results of the reference final review result and the standard final review result) indicates inconsistent compared data, it is determined that the comparison result corresponding to the test case indicates inconsistent compared data. If all the obtained comparison results (including the comparison results of each pair of reference intermediate processing results and standard intermediate processing results, and the comparison results of the reference final review result and the standard final review result) indicate consistent compared data, it is determined that the comparison result corresponding to the test case indicates consistent compared data. When the test case is a black-box case, when the server compares the reference review result corresponding to the test case with the standard review result, it only needs to compare the reference final review result and the standard final review result, and the comparison result obtained through this comparison operation is the comparison result corresponding to the test case.

[0100] Then, the server can count the comparison results corresponding to each of the test cases used in the detection task and determine the proportion of the comparison results indicating inconsistent compared data in all the comparison results. If the proportion exceeds the preset proportion threshold, it is determined that the target review system is abnormal; if the proportion does not exceed the preset proportion threshold, it is determined that the target review system is not abnormal; it should be understood that the above preset proportion threshold can be set according to actual needs, and for example, it can be 10%.

[0101] In this way, through the above method, by detecting the proportion of the comparison results showing inconsistent data in all comparison results, it is possible to more accurately detect whether there is an abnormality in the target audit system. By setting a reasonable preset ratio threshold, it is possible to avoid being too insensitive or too sensitive to the abnormality detection of the target audit system, so as to match the relevant business requirements.

[0102] In a possible implementation manner, the test cases used to participate in the detection task are selected from a use case library; a large number of use cases with rich forms are stored in the use case library, and the standard audit results corresponding to each use case are stored. The use case library can provide test data support for the detection tasks of various audit systems, that is, it supports providing available test cases for the detection tasks of various audit systems.

[0103] See Figure 5 , Figure 5 which is a schematic diagram of the construction process of the use case library provided by the embodiments of the present application. As Figure 5 shown, the use case library can be constructed through the following steps:

[0104] Step 501: Obtain candidate use cases and determine the standard audit results corresponding to the candidate use cases.

[0105] In the embodiments of the present application, the server can obtain a large number of candidate use cases. Specifically, the candidate use cases can be obtained by crawling on the network, or the candidate use cases independently made by relevant staff can be obtained. No limitation is made on the obtaining method of the candidate use cases here. It should be understood that the candidate use cases can be any form of content, such as text, image, video, association relationship (including various contents with association relationships), etc. No limitation is made on the form of the candidate use cases here.

[0106] For each obtained candidate use case, it is also necessary to determine the standard audit result corresponding to the candidate use case. Exemplarily, the standard audit result corresponding to the candidate use case can be determined by manual annotation, or the standard audit result corresponding to the candidate use case can be determined by obtaining historical audit records. For example, when the candidate use case is a typical use case that has been historically audited, the historical audit record generated when auditing the candidate use case can be retrieved, and the standard audit result corresponding to the candidate use case can be obtained from it.

[0107] It should be noted that when determining the corresponding standard review result for a candidate use case, it is necessary to distinguish the type of the candidate use case, that is, to distinguish whether the candidate use case is a black-box use case or a white-box use case. When the candidate use case is a black-box use case, the standard final review result of the candidate use case can be directly determined; when the candidate use case is a white-box use case, it is necessary to determine at least one standard intermediate processing result and the standard final review result corresponding to the candidate use case. When manually annotating the standard intermediate processing result, it is necessary to fully understand the review logic of the review system that can review the candidate use case in advance, so as to annotate the standard intermediate processing results that each intermediate processing node should obtain for the candidate use case. When retrieving the standard intermediate processing result from the historical review record, the standard intermediate processing results of each intermediate processing node in the review system for the candidate use case can be directly extracted from the historical review record.

[0108] Step 502: Perform mutation processing on the candidate use case to obtain a mutated candidate use case corresponding to the candidate use case; determine the standard review result corresponding to the candidate use case as the standard review result corresponding to the mutated candidate use case.

[0109] After obtaining the candidate use case, it is necessary to perform mutation processing on the candidate use case to obtain a mutated candidate use case corresponding to the candidate use case. It should be understood that the mutation processing here can also be called adversarial processing, that is, the candidate use case itself is deformed to obtain a mutated candidate use case with a different form but the same essence as the candidate use case.

[0110] In practical applications, in order to make the bad content escape the review of the review system, users who publish bad content usually perform mutation processing on the bad content, making it difficult for the review system to effectively identify it; to address this situation and enable the review system to still have a reliable recognition ability for the mutated bad content, the embodiments of the present application will mutate the candidate use case, so that a large number of use cases obtained after mutation processing are stored in the use case library. Correspondingly, using such use cases to detect the review system can effectively detect whether the review ability of the review system for such mutated use cases is reliable.

[0111] In a possible implementation manner, the server can perform mutation processing on the candidate use case in the following way to obtain a mutated candidate use case corresponding to the candidate use case:

[0112] Perform mutation processing on the candidate use case according to a preset mutation coefficient to obtain a mutated candidate use case corresponding to the candidate use case; where the mutation coefficient is used to determine the degree of mutation corresponding to the mutation processing; the mutation processing includes at least one of screen mirroring processing, screen smearing processing, audio voice-changing processing, noise interference processing, acceleration processing, adding a screen border processing, picture-in-picture processing, reshooting processing, resolution change processing, screen cropping processing, grayscale transformation processing, adding a watermark processing, and color saturation change processing.

[0113] Specifically, a coefficient of variation α can be preset. This coefficient of variation is used to determine the degree of variation corresponding to the mutation processing performed on the candidate use cases. It should be understood that the larger the coefficient of variation, the higher the degree of variation corresponding to the mutation processing performed.

[0114] When performing mutation processing on the candidate use cases, at least one of the following processing methods can be executed: screen mirroring processing, screen smearing processing, audio voice-changing processing, noise interference processing, acceleration processing, adding a screen border processing, picture-in-picture processing, reshooting processing, resolution change processing, screen cropping processing, grayscale transformation processing, adding a watermark processing, and color saturation change processing, to obtain corresponding mutated candidate use cases.

[0115] Among them, screen mirroring processing refers to mirroring the screen of the candidate use case in the form of an image or video; screen smearing processing refers to smearing a local area of the screen of the candidate use case in the form of an image or video; audio voice-changing processing refers to changing the voice of the audio in the candidate use case in the form of a video; noise interference processing refers to adding noise interference to the candidate use case in the form of an image or video, such as adding noise interference to the screen or adding noise interference to the audio; acceleration processing refers to changing the original playback speed of the candidate use case in the form of a video; adding a screen border processing refers to adding a screen border to the original screen of the candidate use case in the form of an image or video; picture-in-picture processing refers to adding a sub-screen to the original screen of the candidate use case in the form of an image or video; reshooting processing refers to reshooting the candidate use case in the form of an image or video; resolution change processing refers to changing the original resolution of the candidate use case in the form of an image or video; screen cropping processing refers to cropping the screen in the candidate use case in the form of an image or video; grayscale transformation processing refers to changing the original grayscale of the candidate use case in the form of an image or video; adding a watermark processing refers to adding a watermark to the original screen of the candidate use case in the form of an image or video; color saturation change processing refers to changing the original color saturation of the candidate use case in the form of an image or video.

[0116] It should be understood that when specifically performing mutation processing on the candidate use cases, one or more of the above mutation processing methods can be randomly selected for execution, or one or more specific mutation processing methods can be selected according to actual needs for execution. The embodiments of the present application do not make any limitations on the specific mutation processing methods used here.

[0117] In this way, by mutating the candidate test cases in the above manner, the generalization of the test cases stored in the test case library can be improved, making the test cases stored in the test case library more diverse. Moreover, using the mutated candidate test cases obtained through mutation processing to detect the target audit system can more accurately detect whether the audit ability of the target audit system is reliable.

[0118] After mutating the candidate test cases to obtain the corresponding mutated candidate test cases, the standard audit result corresponding to the candidate test case can be directly determined as the standard audit result corresponding to the mutated candidate test case. That is, mutating the candidate test cases only changes the manifestation form of the candidate test case, without changing its essential content. Therefore, the mutated candidate test cases obtained through mutation processing should correspond to the same standard audit result as the candidate test case.

[0119] Step 503: Enter the candidate test case and its corresponding standard audit result, as well as the mutated candidate test case and its corresponding standard audit result into the test case library.

[0120] Furthermore, enter the above candidate test case and its corresponding standard audit result, as well as the mutated candidate test case and its corresponding standard audit result into the test case library in the embodiment of the present application.

[0121] In this way, by constructing the test case library in the above manner, it can be ensured that the test cases stored in the test case library are diverse, and at the same time, it can be ensured that the test cases stored in the test case library have high reference value for the detection task of the audit system. That is, using the test cases stored in the test case library in the detection task is beneficial to accurately detecting whether there is an abnormality in the corresponding audit system, providing reliable data support for the detection task of the audit system.

[0122] In a possible implementation manner, the task configuration information corresponding to the detection task includes the call interface information of the target audit system to be detected, and the call interface information is used to indicate the call interface of the target audit system. Correspondingly, "invoking the target system to process the test case to obtain the reference processing result corresponding to the test case" in the above step 203 includes:

[0123] Invoking the infrastructure layer that supports the operation of the target audit system to audit the test case to obtain the original reference audit result corresponding to the test case;

[0124] Through the proxy module, obtaining the original reference audit result corresponding to the test case generated in the infrastructure layer;

[0125] Performing standardization processing on the original reference audit result to obtain the reference audit result.

[0126] Specifically, in practical applications, the operation of the target audit system depends on the infrastructure layer, which can provide software and hardware support related to the operation of the target audit system, such as storage services, computing engines, network services, middleware, operating systems, etc.

[0127] Correspondingly, in the embodiments of the present application, the server can determine the call interface of the infrastructure layer that supports the operation of the target audit system based on the call interface information included in the task configuration information, and then call the infrastructure layer through this call interface to support the operation of the target audit system, so as to audit the determined test cases and obtain the original reference audit result determined by the target audit system for this test case. It should be understood that the original reference audit result here is the audit result originally output by the target audit system for the test case. Generally, the data format of the original reference audit result is different from the data format of the standard audit result. In order to facilitate subsequent comparison with the standard audit result, further processing of the original reference audit result is required.

[0128] Since the infrastructure layer usually does not actively report the original reference audit result to the server that manages the execution of the entire detection task, the server needs to pull the generated original reference audit result from the infrastructure layer through the proxy module. As introduced above, since the data format of the original reference audit result is usually different from the data format of the standard audit result, in order to facilitate subsequent comparison of the two, the server also needs to perform standardization processing on the original reference audit result; specifically, the server can implement the standardization processing of the original reference audit result through data warehouse technology (Extract-Transform-Load, ETL) to obtain a reference audit result with the same data format as the standard audit result, and store this reference audit result.

[0129] In this way, through the above method of interacting with the infrastructure layer that supports the target audit system, controlling the target audit system to audit the test cases, obtaining the original reference audit result, and performing standardization processing on the original reference audit result to obtain a reference audit result with the same data format as the standard audit result, it is convenient for subsequent comparison of the two, providing a more standardized implementation process for the detection task, and thus facilitating the efficient execution of the detection task.

[0130] In a possible implementation manner, the task configuration information corresponding to the detection task includes the detection concurrency number of the detection task, and the detection concurrency number is used to indicate the number of test cases audited by calling the target audit system within a single time unit. Correspondingly, the above "call the infrastructure layer that supports the operation of the target audit system to audit the test cases and obtain the original reference audit result corresponding to the test cases" includes:

[0131] In each time unit, send the test cases to the infrastructure layer according to the number indicated by the detected concurrency number, so that the infrastructure layer audits the received test cases to obtain the original reference audit results corresponding to the test cases.

[0132] Exemplarily, the detected concurrency number can be used to indicate that 100 test cases are put into the target audit system per hour, that is, to control the target audit system to audit 100 test cases per hour; correspondingly, the server can put 100 test cases into the infrastructure layer per hour, so that the target audit system in the infrastructure layer audits these 100 test cases to obtain the original reference audit results corresponding to each of the 100 test cases.

[0133] It should be understood that the above detected concurrency number can be flexibly set according to the audit mechanism of the target audit system; for example, if the audit mechanism of the target audit system is relatively complex and it takes more processing resources to complete an audit operation, then the detected concurrency number can be set relatively small accordingly to reduce the occupation of the processing resources of the target audit system by the detection task, avoid excessive waste of processing resources, and at the same time avoid the detection task from affecting the normal business of the target audit system; on the contrary, if the audit mechanism of the target audit system is relatively simple and it does not take much processing resources to complete an audit operation, then the detected concurrency number can be set relatively large accordingly, so that the target audit system audits more test cases, thereby realizing the accurate detection of the target audit system.

[0134] In this way, by setting the detected concurrency number in the above manner and controlling the resource occupation of the detection task on the target audit system based on the detected concurrency number, the accurate detection of the target audit system can be realized without affecting the normal business of the target audit system and without causing waste of processing resources.

[0135] In a possible implementation manner, the task configuration information corresponding to the detection task includes task alarm information, and the task alarm information is used to indicate the alarm method, the alarm object, and the alarm content template. Correspondingly, the method provided in the embodiments of the present application further includes:

[0136] When the detection result corresponding to the target audit system indicates that the target audit system is abnormal, generate target alarm content based on the alarm content template;

[0137] Send the target alarm content to the alarm object according to the alarm method.

[0138] Specifically, when the detection result corresponding to the target audit system indicates that the target audit system is abnormal, the server can generate corresponding target alarm content according to the alarm content template indicated in the task configuration information. For example, assuming the alarm content template is "The target audit system is abnormal, please handle it in time", the target alarm content "The target audit system is abnormal, please handle it in time" can be generated accordingly.

[0139] Furthermore, according to the alarm method indicated in the task alarm information in the task configuration information, such as telephone, text message, email, instant messaging application (such as a certain communication account or a certain communication group in the instant messaging application), etc., the above-mentioned target alarm content is sent to the alarm object indicated in the task alarm information to prompt the alarm object that the target audit system is abnormal.

[0140] In this way, through the above method, when it is detected that the target audit system is abnormal, the corresponding responsible person can be notified in a timely and accurate manner, which helps the responsible person to maintain and adjust the target audit system in a timely manner, so that the target audit system can restore its normal audit ability, avoid the situation of large-scale spread of bad content, and have a negative impact on the network environment and user experience.

[0141] In a possible implementation manner, the method provided by the embodiment of the present application may further include:

[0142] Display the detection result corresponding to the target audit system and the execution information of the detection task; wherein, the detection result is used to indicate whether the target audit system is abnormal, and in the case where the target audit system is abnormal, the detection result is further used to indicate the reason for the abnormality of the target audit system; the execution information includes at least one of the following information: the number of test cases used in the detection task, the number of test cases corresponding to the comparison result indicating that the compared data is inconsistent, the type of test cases used in the detection task, the distribution of different types of test cases used in the detection task, and the distribution of comparison results corresponding to different types of test cases.

[0143] After the server completes the detection task, it can display the detection results corresponding to the target audit system through a specific terminal device (for example, the terminal device facing the target object that initiated the detection task). For example, it can display whether the audit ability of the target audit system obtained by the detection is normal. At the same time, it can also display the relevant execution information of the detection task to reflect the overall execution situation of the detection task. The execution information displayed here can include at least one of the following: the number of test cases used in the detection task, the number of test cases with inconsistent data in the comparison result representation corresponding to the detection task, the type of test cases used in the detection task, the distribution of different types of test cases (such as the respective proportions of various test cases used), and the distribution of comparison results corresponding to different types of test cases.

[0144] In this way, by the above method, displaying the detection results corresponding to the target audit system and the execution information of the detection task helps relevant personnel to more intuitively understand the detection task obtained through the detection task, and at the same time more intuitively understand the situation of test cases used in the entire detection task and the distribution of comparison results corresponding to different types of test cases, which helps relevant personnel to perform targeted operation and maintenance on the target audit system accordingly.

[0145] To facilitate a further understanding of the system detection method provided by the embodiments of the present application, the following combines Figure 6 the implementation technical architecture of the system detection method shown, and gives an overall exemplary introduction to the system detection method. As Figure 6 shown, the implementation technical architecture of the system detection method includes an interaction layer, an application support layer, and an infrastructure layer.

[0146] Among them, the interaction layer includes a Web portal, a terminal application (Application, APP), and an interface platform. The Web portal and the terminal APP are used to support an object to initiate a detection task for the target audit system, select the test cases used in the detection task, and set the task configuration information corresponding to the detection task. The interface platform includes a telephone, a short message, an email, and an instant messaging application, that is, when an alarm message needs to be sent, the alarm message can be sent to the corresponding alarm object through various methods supported by the interface platform.

[0147] Among them, the application support layer includes a dial test function, a use case library, and general functions.

[0148] The probing function is the core of the system detection method provided in the embodiments of the present application. It runs through the core process of the entire detection task. By creating a detection task, selecting test cases, configuring task information, executing the probing engine, collecting and preprocessing probing data, comparing data results, alarming, and the data BI (Business Intelligence) module, the implementation process of the entire detection task is connected in series.

[0149] Figure 7 It is a schematic diagram of the process of the probing function provided in the embodiments of the present application. As Figure 7 shown, in this probing function, a detection task can be created. Specifically, the detection task can be created by responding to the operation of initiating a detection task triggered by an object through the Web portal or the terminal APP. Then, test cases are selected. Specifically, the object initiating the detection task can select the test cases participating in the detection task from the use case library. Next, task information is configured. Specifically, the object initiating the detection task can set the task configuration information corresponding to the detection task through the Web portal or the terminal APP, including but not limited to the execution frequency of the detection task (one-time task or periodic task), task alarm information (indicating the alarm object, alarm method, alarm content template), detection concurrency (indicating the number of test cases launched within a single time unit), task scenario, etc. Furthermore, when the detection task starts, the probing engine can call the infrastructure layer to execute the detection task according to the task configuration information of the detection task to implement the review of the selected test cases. After completing the review of the test cases, through probing data collection and preprocessing, the Agent is used to pull the obtained original reference review results from the infrastructure layer, and ETL is used to preprocess the original reference review results to obtain the reference review results. Finally, through data result comparison, the reference review results of the test cases are compared with the standard review results of the test cases entered in the use case library. When it is found through comparison that the detected review system cannot accurately review a large number of test cases, alarm information is sent to the corresponding alarm object through the pre-set alarm method. In addition, the data BI module can display the detection results of the detection task, for example, it can display the number of test cases used in the detection task, the number of test cases corresponding to the comparison results indicating inconsistent data, the types of test cases used, the reasons for the anomalies of the review system, etc.

[0150] The use case library is used to manage the use cases that can be put into the detection task and the corresponding standard review results. It mainly includes links such as use case collection, use case mutation, and use case entry.

[0151] Figure 8 It is a schematic diagram of the construction process of the use case library provided in the embodiments of the present application. As Figure 8As shown, in this building process, a large number of candidate test cases can be collected first, and then the types of candidate test cases are distinguished to mark the standard review results. For example, for black-box test cases, the corresponding standard final review results can be directly marked. For white-box test cases, the corresponding intermediate processing results of each standard need to be marked according to the template. In addition, the collected candidate test cases also need to be mutated. For example, according to the preset mutation coefficient, at least one of the following mutation processes is performed on the candidate test cases: screen mirroring, screen smearing, audio voice conversion, noise interference, acceleration, adding a screen border, picture-in-picture, reshooting, resolution change, screen cropping, grayscale transformation, adding a watermark, color saturation change, etc., to obtain mutated candidate test cases, and the standard review results corresponding to the candidate test cases are assigned to the mutated candidate test cases. Finally, the above-mentioned candidate test cases and their corresponding standard review results, as well as the mutated candidate test cases and their corresponding standard review results, are all entered into the test case library.

[0152] The general function is used to provide support for scheduling management capabilities and data analysis and processing capabilities for the execution of the detection task through ETL, proxy modules, report BI modules, etc.

[0153] The infrastructure layer is used to support the operation of the audit system to be detected and provide software and hardware support such as storage services, computing engines, network services, middleware, and operating systems for it.

[0154] For the audit system detection method described above, the present application also provides a corresponding system detection device to enable the above system detection method to be applied and implemented in practice.

[0155] See Figure 9 , Figure 9 is a schematic structural diagram of a system detection device 900 corresponding to the system detection method shown above. As Figure 2 shown, the system detection device 900 includes: Figure 9 shown, the system detection device 900 includes:

[0156] A task creation module 901, configured to create a detection task for a target system;

[0157] An information determination module 902, configured to determine test cases for participating in the detection task, and determine the standard processing results corresponding to the test cases; determine the task configuration information corresponding to the detection task, where the task configuration information is used to indicate relevant execution attributes of the detection task;

[0158] The task execution module 903 is configured to execute the detection task according to the task configuration information when the detection task meets the task start condition; in the detection task, call the target system to process the test case, obtain the reference processing result corresponding to the test case, compare the reference processing result corresponding to the test case with the standard processing result, and determine the detection result corresponding to the target system according to the comparison result.

[0159] Optionally, the target system is a target audit system; the test case is a white-box case, and the standard processing result corresponding to the white-box case includes a standard intermediate processing result and a standard final audit result, and the standard intermediate processing result corresponds to an intermediate processing node in the audit system; then the task execution module 903 is specifically configured to:

[0160] Call the target audit system to audit the test case, obtain the reference intermediate processing result determined by the intermediate processing node in the target audit system for the test case, and the reference final audit result finally output by the target audit system for the test case;

[0161] Compare whether the reference intermediate processing result and the standard intermediate processing result with a corresponding relationship are consistent to obtain an intermediate comparison result; and, compare whether the reference final audit result and the standard final audit result are consistent to obtain a final comparison result;

[0162] Determine whether there is an abnormality in the target audit system according to the intermediate comparison result and the final comparison result; in the case of determining that the target audit system has an abnormality, determine the abnormal intermediate processing node in the target audit system according to the intermediate comparison result.

[0163] Optionally, the target system is a target audit system; the test case is a black-box case, and the standard processing result corresponding to the black-box case is a standard final audit result; then the task execution module 903 is specifically configured to:

[0164] Call the target audit system to audit the test case, obtain the reference final audit result finally output by the target audit system for the test case;

[0165] Compare whether the reference final audit result corresponding to the test case is consistent with the standard final audit result, and determine the detection result corresponding to the target audit system according to the comparison result.

[0166] Optionally, if there are multiple test cases participating in the detection task, the task execution module 903 is specifically configured to:

[0167] For each of the test cases, compare whether the reference processing result corresponding to the test case is consistent with the standard processing result to obtain the comparison result corresponding to the test case;

[0168] Count the comparison results corresponding to each of the test cases, and determine the proportion of the comparison results indicating inconsistent compared data in all the comparison results;

[0169] When the proportion exceeds a preset proportion threshold, determine that there is an abnormality in the target system.

[0170] Optionally, the test cases are selected from a use case library; the device further includes a use case library construction module, and the use case library construction module includes:

[0171] A use case acquisition unit, configured to acquire candidate use cases and determine the standard processing results corresponding to the candidate use cases;

[0172] A use case mutation unit, configured to perform mutation processing on the candidate use cases to obtain mutated candidate use cases corresponding to the candidate use cases; determine the standard processing results corresponding to the candidate use cases as the standard processing results corresponding to the mutated candidate use cases;

[0173] A use case entry unit, configured to enter the candidate use cases and their corresponding standard processing results, as well as the mutated candidate use cases and their corresponding standard processing results, into the use case library.

[0174] Optionally, the use case mutation unit is specifically configured to:

[0175] Perform mutation processing on the candidate use cases according to a preset mutation coefficient to obtain mutated candidate use cases corresponding to the candidate use cases;

[0176] Wherein, the mutation coefficient is used to determine the degree of mutation corresponding to the mutation processing; the mutation processing includes at least one of screen mirroring processing, screen smearing processing, audio voice change processing, noise interference processing, acceleration processing, adding a screen border processing, picture-in-picture processing, reshooting processing, resolution change processing, screen cropping processing, grayscale transformation processing, adding a watermark processing, and color saturation change processing.

[0177] Optionally, the target system is a target audit system; the task execution module 903 is specifically configured to:

[0178] Call the infrastructure layer that supports the operation of the target audit system to audit the test cases to obtain the original reference audit results corresponding to the test cases;

[0179] Through the proxy module, obtain the original reference audit results corresponding to the test cases generated in the infrastructure layer;

[0180] Standardize the original reference review result to obtain the reference review result as the reference processing result.

[0181] Optionally, the task configuration information includes the detection concurrency number of the detection task, and the detection concurrency number is used to indicate the number of test cases for auditing the target audit system called within a single time unit; then the task execution module 903 is specifically configured to:

[0182] Within each time unit, send the test cases to the infrastructure layer according to the number indicated by the detection concurrency number, so that the infrastructure layer audits the received test cases to obtain the original reference review result corresponding to the test cases.

[0183] Optionally, the task configuration information includes task warning information, and the task warning information is used to indicate the warning method, warning object, and warning content template; then the task execution module 903 is further configured to:

[0184] When the detection result corresponding to the target system indicates that the target system is abnormal, generate target warning content based on the warning content template;

[0185] Send the target warning content to the warning object according to the warning method.

[0186] Optionally, the task execution module 903 is further configured to:

[0187] Display the detection result corresponding to the target system and the execution information of the detection task;

[0188] Wherein, the detection result is used to indicate whether the target system is abnormal, and in the case where the target system is abnormal, the detection result is further used to indicate the reason for the abnormality of the target system; the execution information includes at least one of the following information: the number of test cases used in the detection task, the number of test cases corresponding to the comparison result indicating that the compared data is inconsistent, the type of test cases used in the detection task, the distribution of different types of test cases used in the detection task, and the distribution of comparison results corresponding to different types of test cases.

[0189] The above device proposes a dedicated automated detection mechanism for the target system to autonomously detect whether there are abnormalities in the target system, that is, to detect whether the target system can accurately identify inappropriate content. This method creates a detection task, uses the target system to be detected to review test cases with pre-labeled standard processing results, obtains reference processing results, and then determines the reliability of the current review ability of the target system by comparing whether the reference processing results are consistent with the standard processing results, that is, judges whether the target system can accurately distinguish normal content from inappropriate content at present. Thus, the actual review of real test cases is carried out using the target system to achieve accurate detection of the review ability of the target system. In addition, the device can automatically execute the detection task without any operation by relevant staff in the detection task, which can save relevant resources and improve the detection efficiency. Correspondingly, operating and maintaining the target system based on the detection results obtained through the above device can more reliably maintain a healthy network environment, thereby providing a better Internet experience for users.

[0190] The embodiment of the present application also provides a computer device for a detection system. This computer device can specifically be a terminal device or a server. The terminal device and the server provided by the embodiment of the present application will be introduced from the perspective of hardware implementation below.

[0191] See Figure 10 , Figure 10 is a schematic structural diagram of the terminal device provided by the embodiment of the present application. As Figure 10 shown, for the sake of convenience of description, only parts related to the embodiment of the present application are shown. For specific technical details not disclosed, please refer to the method part of the embodiment of the present application. The terminal can be any terminal device including a mobile phone, a tablet computer, a personal digital assistant (PDA), a point of sales (POS), an in-vehicle computer, etc. Taking the terminal as a computer as an example:

[0192] Figure 10 What is shown is a block diagram of a part of the structure of a computer related to the terminal provided by the embodiment of the present application. Referring to Figure 10 , the computer includes: a radio frequency (RF) circuit 1010, a memory 1020, an input unit 1030 (including a touch panel 1031 and other input devices 1032), a display unit 1040 (including a display panel 1041), a sensor 1050, an audio circuit 1060 (which can be connected to a speaker 1061 and a microphone 1062), a wireless fidelity (WiFi) module 1070, a processor 1080, and a power supply 1090, etc. Those skilled in the art can understand,Figure 10 The computer architecture shown does not constitute a limitation on the computer, and may include more or fewer components than shown, or combine certain components, or have a different component arrangement.

[0193] The memory 1020 can be used to store software programs and modules. The processor 1080 executes various functional applications and data processing of the computer by running the software programs and modules stored in the memory 1020. The memory 1020 mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the computer (such as audio data, phone book, etc.). In addition, the memory 1020 can include high-speed random access memory, and can also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other non-volatile solid-state storage devices.

[0194] The processor 1080 is the control center of the computer, connecting various parts of the entire computer using various interfaces and lines. By running or executing the software programs and / or modules stored in the memory 1020, and by calling the data stored in the memory 1020, it executes various functions of the computer and processes data. Optionally, the processor 1080 can include one or more processing units; preferably, the processor 1080 can integrate an application processor and a modem processor. Among them, the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communication. It can be understood that the above-mentioned modem processor may not be integrated into the processor 1080.

[0195] In the embodiments of the present application, the processor 1080 included in the terminal is further used to execute the steps of any implementation manner of the system detection method provided in the embodiments of the present application.

[0196] See Figure 11 , Figure 11Schematic diagram of a server 1100 provided by an embodiment of the present application. The server 1100 may vary greatly due to configuration or performance differences, and may include one or more central processing units (CPUs) 1122 (for example, one or more processors) and a memory 1132, and one or more storage media 1130 (for example, one or more mass storage devices) for storing application programs 1142 or data 1144. Among them, the memory 1132 and the storage media 1130 may be transient storage or persistent storage. The program stored in the storage media 1130 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations on the server. Further, the central processing unit 1122 may be configured to communicate with the storage media 1130 and execute a series of instruction operations in the storage media 1130 on the server 1100.

[0197] The server 1100 may further include one or more power supplies 1126, one or more wired or wireless network interfaces 1150, one or more input / output interfaces 1158, and / or one or more operating systems, such as Windows Server TM , Mac OS X TM , Unix TM , Linux TM , FreeBSD TM and so on.

[0198] The steps performed by the server in the above embodiments may be based on the Figure 11 server structure shown.

[0199] Among them, the CPU 1122 may also be used to execute the steps of any implementation manner of the system detection method provided by the embodiment of the present application.

[0200] The embodiment of the present application further provides a computer-readable storage medium for storing a computer program, and the computer program is used to execute any implementation manner of the system detection method described in the foregoing embodiments.

[0201] The embodiment of the present application further provides a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes any implementation manner of the system detection method described in the foregoing embodiments.

[0202] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0203] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0204] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0205] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

[0206] If the above-mentioned integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs that can store computer programs.

[0207] It should be understood that in this application, "at least one (item)" means one or more, and "a plurality" means two or more. "And / or" is used to describe the relationship between associated objects and indicates that there can be three relationships. For example, "A and / or B" can mean: only A exists, only B exists, and both A and B exist simultaneously. Here, A and B can be singular or plural. The character " / " generally indicates an "or" relationship between the associated objects before and after. "At least one (of the following)" or a similar expression refers to any combination of these items, including any combination of single items or plural items. For example, at least one (of) a, b, or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a, b, and c", where a, b, and c can be single or multiple.

[0208] As described above, the above embodiments are only used to illustrate the technical solutions of this application and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments or perform equivalent replacements for some of the technical features. These modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A system detection method, characterized in that, the method includes: creating a detection task for a target system; determining test cases to participate in the detection task, and determining the standard processing results corresponding to the test cases; determining the task configuration information corresponding to the detection task, where the task configuration information is used to indicate the relevant execution attributes of the detection task; when the detection task meets the task start condition, execute the detection task according to the task configuration information; in the detection task, call the target system to process the test cases, obtain the reference processing results corresponding to the test cases, compare the reference processing results corresponding to the test cases with the standard processing results, and determine the detection results corresponding to the target system according to the comparison results.

2. The method according to claim 1, characterized in that, the target system is a target audit system; the test cases are white-box cases, and the standard processing results corresponding to the white-box cases include standard intermediate processing results and standard final audit results, and the standard intermediate processing results correspond to the intermediate processing nodes in the audit system; the calling the target system to process the test cases to obtain the reference processing results corresponding to the test cases includes: calling the target audit system to audit the test cases to obtain the reference intermediate processing results determined by the intermediate processing nodes in the target audit system for the test cases and the reference final audit results finally output by the target audit system for the test cases.

3. The method according to claim 2, characterized in that, the comparing the reference processing results corresponding to the test cases with the standard processing results and determining the detection results corresponding to the target system according to the comparison results includes: comparing whether the reference intermediate processing results and the standard intermediate processing results with corresponding relationships are consistent to obtain an intermediate comparison result; and comparing whether the reference final audit results and the standard final audit results are consistent to obtain a final comparison result; determining whether there is an abnormality in the target audit system according to the intermediate comparison result and the final comparison result; in the case of determining that there is an abnormality in the target audit system, determining the abnormal intermediate processing node in the target audit system according to the intermediate comparison result.

4. The method according to claim 1, characterized in that, the target system is a target audit system; the test cases are black-box cases, and the standard audit result corresponding to the black-box cases is the standard final audit result; the calling the target system to process the test cases to obtain the reference processing results corresponding to the test cases includes: calling the target audit system to audit the test cases to obtain the reference final audit results finally output by the target audit system for the test cases; the comparing the reference processing results corresponding to the test cases with the standard processing results and determining the detection results corresponding to the target audit system according to the comparison results includes: Compare whether the reference final review result corresponding to the test case is consistent with the standard final review result, and determine the detection result corresponding to the target review system according to the comparison result.

5. The method according to any one of claims 1 to 4, wherein, there are multiple test cases for participating in the detection task, and comparing the reference processing result corresponding to the test case with the standard processing result, and determining the detection result corresponding to the target system according to the comparison result includes: For each test case, compare whether the reference processing result corresponding to the test case is consistent with the standard processing result to obtain the comparison result corresponding to the test case; Count the comparison results corresponding to each test case, and determine the proportion of the comparison results indicating inconsistent compared data in all the comparison results; When the proportion exceeds a preset proportion threshold, it is determined that the target system is abnormal.

6. The method according to any one of claims 1 to 4, wherein, the test case is selected from a use case library; the use case library is constructed in the following manner: Obtain candidate use cases and determine the standard processing results corresponding to the candidate use cases; Perform mutation processing on the candidate use cases to obtain mutated candidate use cases corresponding to the candidate use cases; Determine the standard processing result corresponding to the candidate use case as the standard processing result corresponding to the mutated candidate use case; Enter the candidate use case and its corresponding standard processing result, as well as the mutated candidate use case and its corresponding standard processing result into the use case library.

7. The method according to claim 6, wherein, the performing mutation processing on the candidate use cases to obtain mutated candidate use cases corresponding to the candidate use cases includes: Performing mutation processing on the candidate use cases according to a preset mutation coefficient to obtain mutated candidate use cases corresponding to the candidate use cases; wherein, the mutation coefficient is used to determine the degree of mutation corresponding to the mutation processing; the mutation processing includes at least one of picture mirroring processing, picture smearing processing, audio voice-changing processing, noise interference processing, acceleration processing, adding a picture border processing, picture-in-picture processing, reshooting processing, resolution change processing, picture cropping processing, grayscale transformation processing, adding a watermark processing, and color saturation change processing.

8. The method according to claim 1, wherein, the target system is a target review system; the calling the target system to process the test case to obtain the reference processing result corresponding to the test case includes: Calling the infrastructure layer that supports the operation of the target review system to review the test case to obtain the original reference review result corresponding to the test case; Through the proxy module, obtain the original reference review result corresponding to the test case generated in the infrastructure layer; Perform standardization processing on the original reference review result to obtain a reference review result as the reference processing result.

9. The method according to claim 8, wherein, The task configuration information includes the detection concurrency of the detection task, and the detection concurrency is used to indicate the number of test cases audited by invoking the target audit system within a single time unit; The invocation supports the infrastructure layer that enables the target audit system to operate to audit the test cases, and obtains the original reference audit results corresponding to the test cases, including: Within each time unit, send the test cases to the infrastructure layer according to the number indicated by the detection concurrency, so that the infrastructure layer audits the received test cases to obtain the original reference audit results corresponding to the test cases.

10. The method according to claim 1 or 8, wherein, The task configuration information includes task warning information, and the task warning information is used to indicate the warning method, warning object and warning content template; the method further includes: When the detection result corresponding to the target system indicates that the target system is abnormal, generate target warning content based on the warning content template; Send the target warning content to the warning object according to the warning method.

11. The method according to claim 1 or 8, wherein, The method further includes: Display the detection result corresponding to the target system and the execution information of the detection task; Among them, the detection result is used to indicate whether the target system is abnormal, and when the target system is abnormal, the detection result is further used to indicate the cause of the abnormality of the target system; the execution information includes at least one of the following information: the number of test cases used in the detection task, the number of test cases corresponding to the comparison result indicating that the compared data is inconsistent, the type of test cases used in the detection task, the distribution of different types of test cases used in the detection task, and the distribution of comparison results corresponding to different types of test cases.

12. A system detection device, wherein, The device includes: A task creation module, configured to create a detection task for a target system; An information determination module, configured to determine the test cases to participate in the detection task, and determine the standard processing results corresponding to the test cases; determine the task configuration information corresponding to the detection task, and the task configuration information is used to indicate the relevant execution attributes of the detection task; A task execution module, configured to execute the detection task according to the task configuration information when the detection task meets the task start condition; in the detection task, invoke the target system to process the test cases to obtain the reference processing results corresponding to the test cases, compare the reference processing results corresponding to the test cases with the standard processing results, and determine the detection result corresponding to the target system according to the comparison result.

13. A computer device, wherein, The device includes a processor and a memory; The memory is used to store a computer program; The processor is configured to execute the system detection method according to any one of claims 1 to 11 according to the computer program.

14. A computer-readable storage medium, characterized in that, the computer-readable storage medium is used to store a computer program, and the computer program is used to execute the system detection method according to any one of claims 1 to 11.

15. A computer program product, comprising a computer program or instructions, characterized in that, when the computer program or the instructions are executed by a processor, the system detection method according to any one of claims 1 to 11 is implemented.

Images