Promotion information testing method and device, storage medium and program product

Abstract

The application discloses a promotion information test method and device, a storage medium and a program product. In response to a test operation instruction, a content page is entered through a target promotion information bit portal, promotion information judgment processing is performed on the content page, a promotion information judgment result is obtained, when the promotion information judgment result is that the content page contains target promotion information, trigger test is performed on the target promotion information, after the trigger test is performed, a first page image is obtained by intercepting screen display content, the first page image is input into a promotion information page detection model, a first detection result is obtained, and then the correctness of the first page image is determined according to the first detection result. The application can realize automatic test of promotion information, and improve test efficiency and test effectiveness of promotion information test. The application can be widely applied in test technology of software in terminal devices such as a vehicle-mounted terminal, a tablet computer, a smart phone, a smart watch, a smart bracelet and the like.

Description

Technical Field

[0001] This application relates to the field of software testing technology, and in particular to a promotional information testing method and apparatus, storage medium, and program product. Background Technology

[0002] With the development of media information technology, users can obtain information through more channels, thereby satisfying their diverse information needs. To cater to users' growing demand for information, promotional information providers often push promotional information to users through various internet platforms and push notification channels.

[0003] To ensure users can effectively access promotional information, it's often necessary to test the promotional information to determine if it redirects correctly. Currently, most promotional information testing is done manually by testers, which is cumbersome, labor-intensive, and involves a lot of repetitive work, leading to low testing efficiency and limited effectiveness. Summary of the Invention

[0004] The following is an overview of the subject matter described in detail herein. This overview is not intended to limit the scope of the claims.

[0005] This application provides a method, apparatus, storage medium, and program product for testing promotional information, which can realize automated testing of promotional information and improve the testing efficiency and effectiveness of promotional information testing.

[0006] On the one hand, embodiments of this application provide a method for testing promotional information, including the following steps:

[0007] In response to the test operation command, users are directed to the content page through the target promotion information entry point;

[0008] The content page is processed to determine promotional information, and the result of the promotional information determination is obtained.

[0009] When the determination result of the promotion information is that the content page contains target promotion information, a trigger test is performed on the target promotion information, wherein the trigger test is used to cause the content page to redirect;

[0010] After the trigger test is executed, the screen display content is captured to obtain the first page image;

[0011] The first page image is input into the promotion information page detection model to obtain the first detection result of the page elements in the first page image, wherein the page elements are information-carrying elements in the page used to display page information;

[0012] The correctness of the first page image is determined based on the first detection result and the page elements of the preset target promotion information page.

[0013] On the other hand, embodiments of this application also provide a promotional information testing device, including:

[0014] The interaction module is used to respond to test operation commands and enter the content page through the target promotion information entry point;

[0015] The promotion information judgment module is used to perform promotion information judgment processing on the content page and obtain the promotion information judgment result;

[0016] The trigger test module is used to perform a trigger test on the target promotion information when the promotion information determination result is that the content page contains target promotion information, wherein the trigger test is used to cause the content page to redirect;

[0017] The first capture module is used to capture the screen display content to obtain the first page image after the trigger test is executed;

[0018] The page detection module is used to input the first page image into the promotion information page detection model to obtain the first detection result of the page elements in the first page image, wherein the page elements are information-carrying elements in the page used to display page information;

[0019] The result determination module is used to determine the correctness of the first page image based on the first detection result and the page elements of the preset target promotion information page.

[0020] Optionally, the result determination module includes:

[0021] The first comparison module is used to compare the first detection result with the page elements of the preset target promotion information page to obtain the first comparison result;

[0022] The correctness determination module is used to determine the correctness of the first page image based on the first comparison result.

[0023] Optionally, the first comparison module includes at least one of the following:

[0024] The first comparison submodule is used to compare the page element type detection information with the preset page element type information of the target promotion information page when the first detection result includes page element type detection information.

[0025] or,

[0026] The second comparison submodule is used to compare the page element layout detection information with the preset page element layout information of the target promotion information page when the first detection result includes page element layout detection information.

[0027] Optionally, the correctness determination module includes:

[0028] The first determining submodule is used to determine that the first page image is correct when the first comparison result is that the first detection result is consistent with the page elements of the target promotion information page;

[0029] or,

[0030] The second determining submodule is used to determine that the first page image is incorrect when the first comparison result is that the first detection result is inconsistent with the page elements of the target promotion information page.

[0031] Optionally, the promotion information judgment module includes:

[0032] The second capture module is used to capture the content page to obtain a second page image;

[0033] The type prediction module is used to input the second page image into the promotion information type prediction model to obtain a second prediction result;

[0034] The information judgment module is used to perform information judgment processing on the first promotional information when it is determined from the second prediction result that the second page image contains the first promotional information, and to obtain the promotional information judgment result.

[0035] Optionally, the test operation instruction includes a promotional information identifier, and the first promotional information includes promotional information text; the information judgment module includes:

[0036] The information recognition module is used to perform information recognition processing on the first promotional information to obtain the promotional information text;

[0037] The first sending module is used to send the promotion information identifier to the server, so that the server can obtain the target text information based on the promotion information identifier;

[0038] The first receiving module is used to receive the target text information sent by the server;

[0039] The information judgment submodule is used to compare the promotional information text with the target text information. When the promotional information text matches the target text information, the first promotional information is processed for element information judgment to obtain the promotional information judgment result.

[0040] Optionally, the information judgment submodule includes:

[0041] The element detection module is used to input the second page image into the element information detection model to obtain the third detection result;

[0042] The second comparison module is used to compare the third detection result with the preset target element information. When the third detection result is consistent with the target element information, a promotion information judgment result is obtained that the first promotion information is the target promotion information.

[0043] Optionally, the second prediction result includes the promotion information prediction type and confidence level; the information judgment module includes:

[0044] The third comparison module is used to compare the confidence level with the confidence threshold to obtain a second comparison result;

[0045] The promotion information determination module is used to determine that the second page image contains first promotion information when the second comparison result is that the confidence level is greater than or equal to the confidence level threshold, and the promotion information type of the first promotion information is consistent with the promotion information prediction type.

[0046] Optionally, the trigger test module includes:

[0047] The region prediction module is used to input the second page image into the trigger region prediction model to obtain the trigger region prediction information of the target promotion information;

[0048] The region determination module is used to determine the trigger region in the target promotion information based on the trigger region prediction information;

[0049] The trigger test submodule is used to perform trigger tests based on the trigger area.

[0050] Optionally, the test operation instructions include a promotional information identifier; the interaction module includes:

[0051] The second sending module is used to send the promotion information identifier to the server in response to the test operation command, so that the server can obtain the target promotion information position information based on the promotion information identifier;

[0052] The second receiving module is used to receive the target promotion information bit information sent by the server;

[0053] Entry determination module, used to determine the entry point of the target promotion information position based on the target promotion information position information;

[0054] The page entry module is used to access the content page through the target promotion information entry point.

[0055] Optionally, the trigger test module is further configured to re-execute the trigger test on the target promotion information;

[0056] The first capture module is also used to capture the screen display content again to obtain a new first page image after re-executing the trigger test;

[0057] The page detection module is also used to input the new first page image into the promotional information page detection model to obtain a new first detection result;

[0058] The result determination module is further configured to determine the correctness of the new first page image based on the new first detection result, until the number of trigger tests reaches a preset threshold.

[0059] Optionally, the promotional information testing device further includes:

[0060] The first tag adding module is used to add a first type of tag to the image on the first page;

[0061] The first feature annotation module is used to perform first feature annotation processing on the first page image according to the first type of label;

[0062] The first model training module is used to train the promotion information page detection model by using the first page image, which has undergone the first feature annotation process, as a training sample of the promotion information page detection model.

[0063] Optionally, the promotional information testing device further includes:

[0064] The second tag adding module is used to add a second type of tag to the image on the second page;

[0065] The second feature annotation module is used to perform second feature annotation processing on the second page image according to the second type of label;

[0066] The second model training module is used to train the promotion information type prediction model by using the second page image, which has undergone the second feature annotation process, as a training sample of the promotion information type prediction model.

[0067] Optionally, the promotional information testing device further includes:

[0068] The third tag adding module is used to add a third type of tag to the image on the second page;

[0069] The third feature annotation module is used to perform third feature annotation processing on the second page image according to the third type of label;

[0070] The third model training module is used to train the element information detection model by using the second page image, which has undergone the third feature annotation process, as a training sample of the element information detection model.

[0071] Optionally, the promotional information testing device further includes:

[0072] The fourth tag adding module is used to add a fourth type of tag to the image on the second page;

[0073] The fourth feature annotation module is used to perform fourth feature annotation processing on the second page image according to the fourth type of label;

[0074] The fourth model training module is used to train the trigger region prediction model by using the second page image, which has undergone the fourth feature annotation process, as a training sample of the trigger region prediction model.

[0075] On the other hand, embodiments of this application also provide a promotional information testing device, including:

[0076] At least one processor;

[0077] At least one memory for storing at least one program;

[0078] The promotional information testing method described above is implemented when at least one of the programs is executed by at least one of the processors.

[0079] On the other hand, embodiments of this application also provide a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement the promotional information testing method described above.

[0080] On the other hand, embodiments of this application also provide a computer program product, including a computer program or computer instructions, the computer program or computer instructions being stored in a computer-readable storage medium, a processor of a computer device reading the computer program or computer instructions from the computer-readable storage medium, and the processor executing the computer program or computer instructions to cause the computer device to perform the promotional information testing method as described above.

[0081] The embodiments of this application include at least the following beneficial effects: In response to a test operation command, the content page is entered through the target promotion information entry point, and then the promotion information of the content page is judged to obtain the promotion information judgment result. When the promotion information judgment result is that the content page contains target promotion information, a trigger test is performed on the target promotion information. After the trigger test is performed, the screen display content is captured to obtain a first page image, and then the first page image is input into the promotion information page detection model to obtain a first detection result. Then, the correctness of the first page image is determined based on the first detection result. Since the entire promotional information testing process is executed in response to test operation commands, manual testing by testers is unnecessary, achieving automated testing of promotional information and improving testing efficiency. Secondly, after confirming that the content page contains the target promotional information, a trigger test is performed on the target promotional information. This means that each trigger test is specifically for the target promotional information, improving the effectiveness of the testing. Furthermore, after executing the trigger test, a correctness test is performed on the screen display content to check if the displayed content is correct, making the promotional information testing more effective. Additionally, since the correctness test of the screen display content is implemented based on the promotional information page detection model, manual test script setting is unnecessary, and manual modification of the test script is not required for different promotional information pages, thus improving the versatility of promotional information testing.

[0082] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the description, claims and drawings. Attached Figure Description

[0083] The accompanying drawings are used to provide a further understanding of the technical solutions of this application and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of this application and do not constitute a limitation on the technical solutions of this application.

[0084] Figure 1 This is a schematic diagram of an implementation environment provided in an embodiment of this application;

[0085] Figure 2 This is a software architecture diagram of a test terminal provided in one embodiment of this application;

[0086] Figure 3 This is a flowchart of a promotional information testing method provided in one embodiment of this application;

[0087] Figure 4This is a schematic diagram of a content page provided in an embodiment of this application;

[0088] Figure 5 This is a schematic diagram of another content page provided in an embodiment of this application;

[0089] Figure 6 yes Figure 3 A flowchart of a specific method for step 160;

[0090] Figure 7 yes Figure 3 A flowchart of a specific method for step 120;

[0091] Figure 8 yes Figure 7 A flowchart of a specific method for step 123 in the middle;

[0092] Figure 9 yes Figure 7 Flowchart of another specific method for step 123;

[0093] Figure 10 yes Figure 9 A flowchart of a specific method for steps 1236 in the middle;

[0094] Figure 11 yes Figure 3 A flowchart of a specific method for step 130;

[0095] Figure 12 yes Figure 3 A flowchart of a specific method for step 110;

[0096] Figure 13 This is a flowchart illustrating a specific example of a promotional information testing method provided in this application;

[0097] Figure 14 This is a schematic diagram of the second page image provided as a specific example of this application;

[0098] Figure 15 This is a schematic diagram of the second page image provided in another specific example of this application;

[0099] Figure 16 This is a schematic diagram of the promotional information provider profile page provided as a specific example in this application;

[0100] Figure 17 This is a flowchart illustrating the promotional information testing method provided in another specific example of this application;

[0101] Figure 18 This is a flowchart illustrating the promotional information testing method provided in another specific example of this application;

[0102] Figure 19 This is a schematic diagram of a promotional information testing device provided in one embodiment of this application;

[0103] Figure 20 This is a schematic diagram of a promotional information testing device provided in another embodiment of this application. Detailed Implementation

[0104] The present application will be further described below with reference to the accompanying drawings and specific embodiments. The described embodiments should not be considered as limitations on the present application, and all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of the present application.

[0105] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0106] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0107] Before providing a further detailed description of the embodiments of this application, the nouns and terms involved in the embodiments of this application will be explained, and the nouns and terms involved in the embodiments of this application shall be interpreted as follows.

[0108] 1) Interactive applications are those that provide users with a user interface (UI) for interactive operations, such as instant messaging applications and information browsing applications.

[0109] 2) Promotional information entry points refer to the redirection portals that lead to content pages displaying promotional information. Interactive applications may include multiple promotional information entry points. For example, some instant messaging applications may include promotional information entry points for social applications, public service platforms, lightweight applications, and lightweight games. Different promotional information entry points may correspond to different promotional information interfaces. Among them, social applications refer to applications that display social friend information in a news feed format; public service platforms refer to service platforms that can provide various service information and news to target audiences; lightweight applications refer to applications that can be used without downloading and installation; and lightweight games refer to games that can be used without downloading and installation.

[0110] 3) The promotional information interface refers to the terminal screen interface used to display promotional information. The position and extent of the promotional information interface on the terminal screen are relatively flexible and can be set according to actual needs. Different promotional information interfaces can have different interface elements.

[0111] 4) Interface elements refer to the information-carrying elements in the promotional information interface used to display promotional information. Promotional information can include images, text, links, avatars, nicknames, etc. Interface elements constitute the interface style of the promotional information interface. The element information of interface elements includes element type and element position information. Different element types carry different types of promotional information, and element position information includes the coordinates and range of the interface element within the promotional information interface.

[0112] 5) Artificial Intelligence (AI) is the theory, methods, technology, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results. In other words, AI is a comprehensive technology within computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can react in a way similar to human intelligence. AI studies the design principles and implementation methods of various intelligent machines, enabling them to have perception, reasoning, and decision-making capabilities. AI technology is a comprehensive discipline involving a wide range of fields, encompassing both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operating / interactive systems, and mechatronics. AI software technologies mainly include computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0113] 6) Machine Learning (ML) is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory. It specifically studies how computers can simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their performance. Machine learning is the core of artificial intelligence and the fundamental way to endow computers with intelligence; its applications span all areas of artificial intelligence. Machine learning and deep learning typically include techniques such as artificial neural networks, belief networks, reinforcement learning, transfer learning, and inductive learning.

[0114] 7) Deep Learning (DL) is a new research direction in the field of machine learning. It is used to learn the inherent patterns and representational layers of sample data. The information gained in this learning process is of great help in interpreting data such as text, images, and sound. Its ultimate goal is to enable machines to have analytical and learning capabilities like humans, and to be able to recognize data such as text, images, and sound.

[0115] 8) Blockchain is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanisms, and encryption algorithms. Essentially, a blockchain is a decentralized database, a chain of data blocks linked using cryptographic methods. Each data block contains information about a batch of network transactions, used to verify the validity of the information (anti-counterfeiting) and generate the next block. A blockchain can include a blockchain underlying platform, a platform product service layer, and an application service layer. The blockchain underlying platform can include processing modules such as user management, basic services, smart contracts, and operation monitoring. The user management module is responsible for managing the identity information of all blockchain participants, including maintaining public and private key generation (account management), key management, and maintaining the correspondence between user real identities and blockchain addresses (access management). Under authorization, it also monitors and audits transactions of certain real identities and provides risk control rule configuration (risk control audit). The basic service module is deployed on all blockchain node devices to verify the validity of business requests. After consensus is reached on valid requests, they are recorded in storage. For a new business request, the basic service first performs interface adaptation parsing and authentication (interface adaptation), and then encrypts the business information through a consensus algorithm (consensus management). The blockchain process involves several layers of data transmission and processing. The first layer, the blockchain service layer, provides basic capabilities and implementation frameworks for typical applications. Developers can define contract logic using a programming language and publish it to the blockchain (contract registration). The second layer provides blockchain-based application services for business stakeholders. The third layer provides application services based on blockchain solutions for use by business participants.

[0116] To cater to users' information needs, promotional information providers often push promotional information to users through various internet platforms and push notification channels, such as advertising information on specific pages of various applications. To ensure users effectively access this information, it's often necessary to test it to determine if the promotional information redirects correctly. Currently, most promotional information testing is done manually by testers, which suffers from drawbacks such as tedious testing processes, high workload, and repetitive tasks, easily leading to low testing efficiency and effectiveness. To address the technical problems associated with manual testing, a promotional information testing method based on Monkey testing has been proposed. Monkey testing sends pseudo-random user event streams (such as key presses, touchscreen inputs, and gesture inputs) to the terminal's operating system, achieving stress testing of the application. It can randomly generate clicks or touches to simulate user actions, thereby stress testing the application. For example, first, locate the promotional information interface on the page, then simulate clicking on interface elements within the promotional information interface, checking if the page redirects correctly. Then, return to the original page and simulate clicking on other clickable interface elements, checking again if the page redirects correctly, and so on. However, the promotional information testing methods in related technologies can only check whether promotional information appears and whether the page can redirect, but cannot check whether the content of the promotional information (such as image content) is correct or whether the redirected page is correct. Furthermore, these methods only test specific pages, which are pre-defined in the test script, such as native pages or WebView pages (WebView is a control in the Android system used to display web pages). When testing promotional information on different pages, the test script needs to be modified accordingly. Therefore, these methods still suffer from low testing efficiency, low testing effectiveness, and low test versatility.

[0117] To automate the testing of promotional information and improve its efficiency, effectiveness, and versatility, this application provides a method, apparatus, computer-readable storage medium, and computer program product for testing promotional information. Upon receiving a test operation instruction, the system first enters the content page through the target promotional information entry point. Then, it performs promotional information judgment processing on the content page to obtain a judgment result. If the judgment result indicates that the content page contains the target promotional information, a trigger test is performed on the target promotional information. After the trigger test, the screen display content is captured to obtain a first page image. This first page image is then input into the promotional information page detection model to obtain a first detection result. Finally, the correctness of the first page image is determined based on the first detection result. Since the entire promotional information testing process is executed in response to test operation commands, manual testing by testers is unnecessary, achieving automated testing of promotional information and improving testing efficiency. Secondly, after confirming that the content page contains the target promotional information, a trigger test is performed on the target promotional information. This means that each trigger test is specifically for the target promotional information, improving the effectiveness of the testing. Furthermore, after executing the trigger test, a correctness test is performed on the screen display content to check if the displayed content is correct, making the promotional information testing more effective. Additionally, since the correctness test of the screen display content is implemented based on the promotional information page detection model, manual test script setting is unnecessary, and manual modification of the test script is not required for different promotional information pages, thus improving the versatility of promotional information testing.

[0118] Figure 1 This is a schematic diagram of an implementation environment provided in an embodiment of this application. (Refer to...) Figure 1 The implementation environment includes an operation terminal 101, a test terminal 102, and a server 103, with the server 103 communicatively connected to both the operation terminal 101 and the test terminal 102. The operation terminal 101, test terminal 102, and server 103 can be nodes in a blockchain; this embodiment does not impose specific limitations on this.

[0119] The operating terminal 101 can be a tablet computer, laptop computer, desktop computer, or other terminal device, but is not limited to these. Optionally, a test client is installed on the operating terminal 101, and the operating terminal 101 can use the test client to click or touch the promotional information (such as advertising information) displayed on the test terminal 102 to perform other tests.

[0120] The test terminal 102 can be a tablet computer, smartphone, vehicle terminal, smartwatch, smart bracelet, or other terminal device, but is not limited to these. The test terminal 102 and the operation terminal 101 can be directly or indirectly connected via wired or wireless communication methods, and this application embodiment does not specifically limit this.

[0121] Server 103 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0122] like Figure 2 As shown, Figure 2 This is the software architecture diagram of test terminal 102. (Refer to...) Figure 2 The software architecture of the test terminal 102 includes a promotion information service unit, an automation service unit, an AI service unit, a notification service unit, and a storage service unit.

[0123] The promotion information service unit can provide basic services for the normal operation of the application as well as promotion information-related services. For example, basic services include basic client services, public service platform services, and lightweight application services. Promotion information-related services include determining the entry point for promotion information, selecting promotion information content, obtaining promotion information content, and rendering and displaying it.

[0124] The automation service unit provides UI automation and test automation services. The UI automation service automatically renders and displays the UI content; the test automation service performs automated testing on test terminal 102. For example, when test terminal 102 is an Android system terminal, the Uiautomator automation framework can be used to process promotional information; when test terminal 102 is an iOS system terminal, the Webdriveragent automation framework can be used to process promotional information. The Uiautomator automation framework is an automation testing tool for Android system terminals, providing an API (Application Programming Interface) library for customizing UI tests. The Webdriveragent automation framework is an open-source framework for end-to-end UI testing. It can call the underlying iOS XCTest framework, which then calls the iOS system API to execute UI instructions.

[0125] The AI ​​service unit is equipped with AI models, such as a promotional information prediction model and a promotional information page detection model. The promotional information prediction model can include a promotional information type prediction model and an element information detection model. The promotional information type prediction model can predict the promotional information on the page, while the element information detection model can identify interface elements within the promotional information interface. Additionally, the promotional information page detection model can identify the content of the page after a redirect and detect any anomalies. Furthermore, the AI ​​service unit can train each AI model using training samples. For example, it can use relevant information obtained during promotional information testing as training samples to train the promotional information prediction model and the promotional information page detection model.

[0126] The notification service unit can issue a report notification when the promotional information test ends, or issue an alarm notification when an anomaly occurs during the promotional information test, for example, by sending a report notification or alarm notification to the testers via email, instant messaging, or other means.

[0127] The storage service unit is able to save relevant information of the test terminal 102 during the promotion information testing process, such as screenshots of the promotion information interface and the promotion information page after the jump.

[0128] The operating terminal 101 has at least the function of performing click or touch tests on the promotional information displayed on the testing terminal 102. For example, it can respond to user operations in the testing client and send operation instructions to the testing terminal 102, enabling the testing terminal 102 to perform click or touch tests on the displayed promotional information. In addition, the operating terminal 101 also has functions such as downloading test results, generating test reports based on the test results, displaying test reports, and providing download paths for test reports. For example, after completing the promotional information test, the operating terminal 101 can download the test results from the server 103 and generate a test report based on the test results. After generating the test report, the operating terminal 101 can display the test report or provide the user with the download path for the test report.

[0129] The test terminal 102 has at least the following functions: entering a content page through a target promotional information entry point according to an operation command sent by the operation terminal 101; performing promotional information judgment processing on the content page; performing click or touch tests on the target promotional information in the content page; and judging the correctness of the page after the content page is redirected. For example, it can respond to a test operation command sent by the operation terminal 101, enter a content page through a target promotional information entry point, and then perform promotional information judgment processing on the content page to obtain a promotional information judgment result. When the promotional information judgment result is that the content page contains target promotional information, a trigger test is performed on the target promotional information. After the trigger test is performed, the screen display content is captured to obtain a first page image, and then the first page image is input into the promotional information page detection model to obtain a first detection result. Then, the correctness of the first page image is determined based on the first detection result. In addition, the test terminal 102 can also save screenshot information such as the first page image and use this screenshot information to train the promotional information page detection model. In addition, after completing the promotional information test, the test terminal 102 can also send the test results to the server 103, so that the server 103 can save these test results so that it can send the test results according to the request of the operation terminal 101.

[0130] Server 103 has at least the functions of receiving and storing test results sent by test terminal 102 and providing test results for download by operation terminal 101. For example, it can respond to an information download request sent by operation terminal 101 and send the test results uploaded by test terminal 102 to operation terminal 101, so that operation terminal 101 can generate a test report based on the test results, display the test report, or provide the user with the download path of the test report. In addition, server 103 also stores content such as Advertising Identity Document (AID), advertising information position information corresponding to AID, and advertising information text corresponding to AID. Server 103 can respond to an information retrieval request including AID sent by operation terminal 101 or test terminal 102 and send the corresponding advertising information position information and advertising information text to operation terminal 101 or test terminal 102.

[0131] In one optional implementation, the operating terminal 101 is a desktop computer with a test client installed, and the test terminal 102 is a smartphone. In response to a user's test operation in the test client, the operating terminal 101 sends a test operation instruction to the test terminal 102, wherein the test operation instruction includes a promotional information identifier. In response to the test operation instruction, the test terminal 102 obtains the corresponding promotional information entry point information from the server 103 based on the promotional information identifier, determines the target promotional information entry point based on the promotional information entry point information, and then enters the content page through the target promotional information entry point. Next, it performs promotional information judgment processing on the content page to obtain a promotional information judgment result. When the promotional information judgment result indicates that the content page contains target promotional information, the test terminal 102 executes a trigger test on the target promotional information. After executing the trigger test, the test terminal 102 captures the screen display content to obtain a first page image, and then... The page image is input into the promotional information page detection model to obtain the first detection result. Then, the correctness of the first page image is determined based on the first detection result. After completing the promotional information test, the test terminal 102 sends the test result to the server 103. In response to receiving the test result sent by the test terminal 102, the server 103 saves the test result. In response to the information download request sent by the operation terminal 101, the server 103 sends the test result to the operation terminal 101, so that the operation terminal 101 can generate a test report based on the test result. In response to receiving the test result sent by the server 103, the operation terminal 101 generates a test report based on these test results. After the operation terminal 101 generates the test report, it can display the test report or provide the user with the download path of the test report.

[0132] In another optional implementation, the operating terminal 101 is a desktop computer with a test client installed, and the test terminal 102 is a smartphone. In response to a user's test operation in the test client, the operating terminal 101 sends a promotional information request to the server 103, wherein the promotional information request includes a promotional information identifier. In response to the promotional information request, the server 103 obtains the corresponding promotional information entry information based on the promotional information identifier and returns the promotional information entry information to the operating terminal 101. In response to receiving the promotional information entry information, the operating terminal 101 sends a test operation instruction to the test terminal 102, including the promotional information entry information. In response to the test operation instruction, the test terminal 102 determines the target promotional information entry based on the promotional information entry information, then enters the content page through the target promotional information entry, and then performs promotional information judgment processing on the content page to obtain a promotional information judgment result. When the promotional information judgment result indicates that the content page contains target promotional information, the test terminal 102... After triggering the test, the test terminal 102 captures the screen display content to obtain a first page image, and then inputs the first page image into the promotional information page detection model to obtain a first detection result. The correctness of the first page image is then determined based on the first detection result. After completing the promotional information test, the test terminal 102 sends the test result to the server 103. In response to receiving the test result from the test terminal 102, the server 103 saves the test result. In response to an information download request sent by the operation terminal 101, the server 103 sends the test result to the operation terminal 101, enabling the operation terminal 101 to generate a test report based on the test result. In response to receiving the test result from the server 103, the operation terminal 101 generates a test report based on these test results. After generating the test report, the operation terminal 101 can display the test report or provide the user with the download path for the test report.

[0133] Figure 3 This is a flowchart illustrating a promotional information testing method provided in an embodiment of this application. In this embodiment, using... Figure 1 The test terminal 102 in the example is used as the execution subject for illustration. (Refer to...) Figure 3 The promotional information testing method includes, but is not limited to, steps 110 to 160.

[0134] Step 110: In response to the test operation command, enter the content page through the target promotion information entry.

[0135] In this step, when a test operation instruction is received from the operating terminal, the user can respond to the instruction and enter the content page through the target promotion information entry point. This allows subsequent steps to perform click tests or touch tests on the promotion information on the content page, thus achieving the purpose of testing the promotion information.

[0136] It should be noted that the target promotional information entry point is a promotional information entry point in an interactive application. In some embodiments, the interactive application may include multiple promotional information entry points. The target promotional information entry point can be randomly selected from these multiple promotional information entry points, or it can be a specific one among these multiple promotional information entry points. This embodiment does not specifically limit this. For example, when the test operation instruction includes a promotional information identifier, the target promotional information entry point is a specific promotional information entry point determined from these multiple promotional information entry points based on the promotional information identifier. In a specific example, suppose an instant messaging application has four types of promotional information entry points: a public service platform promotional information entry point, a lightweight application promotional information entry point, a lightweight game application promotional information entry point, and a social application promotional information entry point. Different types of promotional information entry points may correspond to different numbers of promotional information entry points. Then, one can be randomly selected from these promotional information entry points as the target promotional information entry point, or the corresponding target promotional information entry point can be determined from these promotional information entry points based on the promotional information identifier.

[0137] It's important to note that different promotional information placements will lead to different content pages. For example, for a social media app promotional placement, you can access the social media app page through the instant messaging application's social media app option. If you then refresh the social media app page, as shown... Figure 4 As shown, a promotional information interface 104 for displaying promotional information can be displayed on the social application page. For example, regarding the entry point for promotional information on a public service platform, one can first find the public service platform option on the main page of the instant messaging application, then click on the public service platform option to enter the public service platform main page, and then click on the specific service account on the public service platform main page to enter the specific information page. At this time, as shown... Figure 5 As shown, the promotion information interface 104, which is used to display promotional information, can be displayed on this information page.

[0138] Step 120: Perform promotional information judgment processing on the content page to obtain the promotional information judgment result.

[0139] It should be noted that a content page accessed through the target promotional information entry point does not necessarily contain promotional information. Therefore, in order to effectively perform subsequent promotional information testing operations, this step involves performing a promotional information judgment process on the content page after accessing it through the target promotional information entry point. This judgment result allows subsequent steps to determine whether to perform promotional information testing operations based on the result. For example, if the promotional information judgment result indicates that the content page contains promotional information, then subsequent steps can perform testing operations on the promotional information on the content page; conversely, if the judgment result indicates that the content page does not contain promotional information, then the current promotional information testing operation should be exited.

[0140] It should be noted that there are different ways to obtain the promotional information judgment result by processing the content page, and this embodiment does not limit the specific implementation. For example, when the promotional information is advertising information, Optical Character Recognition (OCR) technology can be used to identify whether the content page contains text information containing the word "advertisement". If the text information containing the word "advertisement" is identified, the result of judging that the content page contains promotional information can be obtained; if the text information not containing the word "advertisement" is identified, the result of judging that the content page does not contain promotional information can be obtained. Alternatively, a screenshot of the content page can be taken first, and then the screenshot image can be input into the promotional information prediction model for prediction processing. If the prediction result indicates that the screenshot image contains promotional information, the result of judging that the content page contains promotional information can be obtained; if the prediction result indicates that the screenshot image does not contain promotional information, the result of judging that the content page does not contain promotional information can be obtained.

[0141] Step 130: When the promotion information determination result is that the content page contains the target promotion information, execute the trigger test on the target promotion information.

[0142] In this step, since the promotion information judgment result was obtained in step 120, when the promotion information judgment result is that the content page contains the target promotion information, it means that the promotion information displayed on the content page is correct. Therefore, a trigger test can be performed on the target promotion information in the content page so that subsequent steps can further check the correctness of the trigger test and make the promotion information test more effective. The trigger test is used to make the content page redirect.

[0143] It should be noted that since the promotional information interface is the terminal screen interface used to display promotional information, if a content page contains the target promotional information, it means that the content page contains a promotional information interface for displaying that target promotional information. Therefore, the test terminal can generate a trigger message within this promotional information interface to simulate a user's click or touch operation, performing a trigger test on the target promotional information to achieve the purpose of testing the promotional information. For example, when the test terminal is an Android system terminal, it can trigger the Uiautomator automation framework to perform promotional information testing on the target promotional information; when the test terminal is an iOS system terminal, it can use the Webdriveragent automation framework to perform promotional information testing on the target promotional information.

[0144] It should be noted that the trigger test performed on the target promotional information can be different test operations, and this embodiment does not specifically limit it. For example, when the trigger information generated by the test terminal simulates a user's click operation, the trigger test performed is a click test; as another example, when the trigger information generated by the test terminal simulates a user's touch operation, the trigger test performed is a touch test.

[0145] Step 140: After executing the trigger test, capture the screen display content to obtain the first page image.

[0146] It should be noted that although a trigger test was performed on the target promotional information in step 130, it cannot be determined whether the content page was redirected or whether the screen display content after the redirection was correct. To solve these problems, in this step, after the trigger test is performed, the screen display content is captured to obtain the first page image, so that subsequent steps can determine whether the content page was redirected or whether the screen display content after the redirection was correct based on the first page image, thereby making the promotional information test more complete and effective.

[0147] Step 150: Input the first page image into the promotion information page detection model to obtain the first detection result of the page elements in the first page image.

[0148] It's important to note that page elements refer to the information-carrying elements on a webpage used to display page information. This information can include images, text, links, avatars, nicknames, etc. Information-carrying elements are those that display information content. The element information of a page element includes its element type and position information. Different element types carry different types of information content; that is, different element types correspond to different information-carrying elements. Element position information includes the page element's coordinates and location within the page.

[0149] In this step, since the first page image was obtained in step 140, it can be input into the promotion information page detection model to obtain the first detection result of the page elements in the first page image. This allows subsequent steps to determine whether the content page has performed a redirect or whether the screen display content after the redirect is correct, thereby making the promotion information test more complete and effective.

[0150] It should be noted that the promotional information page detection model can be implemented using different algorithms, and this embodiment does not impose any specific limitations on it. For example, the promotional information page detection model can be implemented using a conventional deep neural network algorithm, and its first detection result can be a classification value (such as 1 or 0). Subsequent steps can determine the correctness of the first page image based on this classification value. Alternatively, the promotional information page detection model can be implemented using the YOLO (You Only Look Once) algorithm, and its first detection result can be the page style of the first page image. Subsequent steps can determine the correctness of the first page image based on the page style of the first page image and the page style of the pre-saved promotional information page.

[0151] It's important to note that the YOLO algorithm applies a single Convolutional Neural Network (CNN) to the entire image, dividing it into multiple grids and predicting the class probability and object bounding box size for each grid. The output of the YOLO algorithm can include the center coordinates, height, and width of the object bounding box, as well as the object class. Therefore, the YOLO algorithm can be used to detect objects in images. Furthermore, the YOLO algorithm can predict the probability of an object being present within the bounding box. If this probability is greater than or equal to a preset probability threshold, the corresponding object is considered identified; if the probability is less than the preset threshold, the corresponding object is considered not identified. Therefore, the YOLO algorithm can be used for predictive image processing.

[0152] It should be noted that the first page image obtained during the promotional information testing process can also be used as a training sample for the promotional information page detection model, thereby improving the model's detection accuracy. For example, after capturing the screen display content to obtain the first page image, a first type label can be added to the first page image. Then, based on the first type label, the first feature annotation process can be performed on the first page image. This first page image with the first feature annotation process can then be used as a training sample for the promotional information page detection model. Since the first page image obtained during the promotional information testing process can be used as a training sample for the promotional information page detection model, the number and types of training samples for the model can be effectively expanded, which is beneficial to improving the model's detection accuracy.

[0153] Step 160: Determine the correctness of the first page image based on the first detection result and the page elements of the preset target promotion information page.

[0154] In this step, since the first detection result was obtained in step 150, the correctness of the first page image can be determined based on the first detection result and the page elements of the preset target promotion information page, making the promotion information test more complete and effective.

[0155] In an optional implementation, assuming the promotional information page detection model is implemented by a conventional deep neural network algorithm, and its first detection result is a classification value (such as 1 or 0), then when the first detection result is 1, it can be considered that the content displayed on the screen after the jump is a correct promotional information page, thus determining that the first page image is correct, and at the same time, it can also be determined that the content page has performed the jump normally; when the first detection result is 0, it can be considered that the content page has failed to perform the jump normally, or that the content displayed on the screen after the jump is not a correct promotional information page, thus determining that the first page image is incorrect.

[0156] In another optional implementation, the promotional information page detection model is implemented using the YOLO algorithm. Its first output detection result is the page style of the first page image. When the page style of the first page image matches the page style of the pre-saved target promotional information page (which can be determined based on the page elements of the target promotional information page), it can be considered that the content displayed on the screen after the jump is the correct promotional information page, thus confirming that the first page image is correct and that the content page has successfully performed the jump. When the page style of the first page image does not match the page style of the pre-saved promotional information page, it can be considered that the content page has failed to perform the jump correctly, or that the content displayed on the screen after the jump is not the correct promotional information page, thus confirming that the first page image is incorrect.

[0157] In this embodiment, by adopting the promotion information testing method including the preceding steps 110 to 160, after receiving the test operation instruction, the content page is first entered through the target promotion information entry point, and then the promotion information is judged on the content page to obtain the promotion information judgment result. When the promotion information judgment result is that the content page contains the target promotion information, a trigger test is performed on the target promotion information. After the trigger test is performed, the screen display content is captured to obtain the first page image, and then the first page image is input into the promotion information page detection model to obtain the first detection result. Then, the correctness of the first page image is determined based on the first detection result. Since the entire promotional information testing process is executed in response to test operation commands, manual testing by testers is unnecessary, achieving automated testing of promotional information and improving testing efficiency. Secondly, after confirming that the content page contains the target promotional information, a trigger test is performed on the target promotional information. This means that each trigger test is specifically for the target promotional information, improving the effectiveness of the testing. Furthermore, after executing the trigger test, a correctness test is performed on the screen display content to check if the displayed content is correct, making the promotional information testing more effective. Additionally, since the correctness test of the screen display content is implemented based on the promotional information page detection model, manual test script setting is unnecessary, and manual modification of the test script is not required for different promotional information pages, thus improving the versatility of promotional information testing.

[0158] Reference Figure 6 As shown in the embodiment of this application, step 160 is further described. Step 160 may include, but is not limited to, steps 161 and 162.

[0159] Step 161: Compare the first detection result with the page elements of the preset target promotion information page to obtain the first comparison result.

[0160] In this step, since the first detection result was obtained in step 150, the first detection result can be compared with the page elements of the preset target promotion information page to obtain the first comparison result, so that subsequent steps can determine the correctness of the first page image based on the first comparison result, making the promotion information test more complete and effective.

[0161] It's important to note that the target promotional information page is a pre-saved redirect page corresponding to the target promotional information in the content page, saved before performing the promotional information testing operation. For example, the target promotional information in the content page corresponds to a redirect link; when the target promotional information in the content page is clicked, the content page will redirect to the target promotional information page pointed to by that link. Before performing the promotional information testing operation, this target promotional information page can be saved first. During the promotional information testing operation, after completing the trigger test of the promotional information in the content page, the screen display content can be captured to obtain a first page image. This first page image is then input into the promotional information page detection model to obtain a first detection result. This first detection result is then compared with the page elements of the target promotional information page.

[0162] It should be noted that there are different ways to compare the first detection result with the page elements of the preset target promotion information page, and this embodiment does not limit this in any specific way. For example, when the first detection result includes page element type detection information, the page element type detection information can be compared with the page element type information of the preset target promotion information page; when the first detection result includes page element layout detection information, the page element layout detection information can be compared with the page element layout information of the preset target promotion information page; when the first detection result includes both page element type detection information and page element layout detection information, either the page element type detection information or the page element layout detection information can be compared with the preset target promotion information page first. If the comparison result is consistent, then the remaining one of the page element type detection information and the page element layout detection information can be compared with the target promotion information page.

[0163] It should be noted that page element type information includes text type, image type, and video type, among others. Text type refers to elements whose primary content is text, image type refers to elements whose primary content is image, and video type refers to elements whose primary content is video. Page element layout information refers to the layout style information of page elements.

[0164] Step 162: Determine the correctness of the first page image based on the first comparison result.

[0165] In this step, since the first comparison result was obtained in step 161, it is possible to determine whether the first page image is the page image after the jump based on the first comparison result, and to determine whether the content of the first page image is correct, thereby making the promotional information test more complete and effective.

[0166] It should be noted that there are different ways to determine the correctness of the first page image based on the first comparison result, and this embodiment does not specifically limit this. For example, when the first comparison result is that the first detection result is consistent with the page elements of the target promotional information page, the first page image can be determined to be a correct page image; when the first comparison result is that the first detection result is inconsistent with the page elements of the target promotional information page, the first page image can be determined to be an incorrect page image. It should be noted that there are two specific situations where the first page image is an incorrect page image. One is that the content page has not been redirected, that is, the first page image is a screenshot of the content page; the other is that the content page has been redirected, but the content displayed on the screen after the redirect is not the target promotional information page. Therefore, in this embodiment, the first detection result is first compared with the preset page elements of the target promotional information page to obtain the first comparison result, and then the correctness of the first page image is determined based on the first comparison result. This can solve the technical problem in related technologies that it is impossible to determine whether the content page has been redirected or whether the content displayed on the screen after the redirect is correct, thereby making the promotional information test more complete and effective.

[0167] Reference Figure 7 As shown in the embodiment of this application, step 120 is further described. Step 120 may include, but is not limited to, steps 121 to 123.

[0168] Step 121: Capture the content page to obtain the image of the second page.

[0169] It should be noted that after entering the content page through the target promotion information entry, the content page is the current screen display content. Since the current screen display content may be dynamic, it is not conducive to judging and processing promotion information on the content page. Therefore, in this step, we can first capture a second page image of the content page so that subsequent steps can judge and process promotion information based on the second page image.

[0170] Step 122: Input the image of the second page into the promotion information type prediction model to obtain the second prediction result.

[0171] In this step, since the second page image was obtained in step 121, the second page image can be input into the promotion information type prediction model to obtain the second prediction result, so that subsequent steps can obtain the promotion information judgment result based on the second prediction result.

[0172] It should be noted that the promotional information type prediction model can be implemented by different algorithms, and this embodiment does not impose specific limitations on it. For example, the promotional information type prediction model can be implemented by a conventional deep neural network algorithm, and its output second prediction result can be a classification value (such as 1 or 0). When the second prediction result is 1, it can be determined that the second page image contains promotional information corresponding to the target promotional information type; when the second prediction result is 0, it can be determined that the second page image does not contain promotional information or does not contain promotional information corresponding to the target promotional information type. As another example, the promotional information type prediction model can be implemented by the YOLO algorithm, and its output second prediction result can include the promotional information prediction type and confidence level. When the confidence level meets a preset condition, such as when the confidence level is greater than or equal to a confidence level threshold, it can be determined that the second page image contains promotional information, and the promotional information type of the promotional information is consistent with the predicted promotional information type; when the confidence level does not meet the preset condition, such as when the confidence level is less than a confidence level threshold, it can be determined that the second page image does not contain promotional information, or the second page image contains promotional information, but the promotional information type of the promotional information is inconsistent with the predicted promotional information type.

[0173] It should be noted that the predicted promotional information type refers to the types of promotional information that may be contained in the images on the second page. These promotional information types include banner promotions, interstitial promotions, cover promotions, and video stream promotions. Banner promotions are rectangular promotions that span the entire webpage; clicking on this rectangle links to the promotion provider's webpage. Interstitial promotions are those that pop up in half-screen or full-screen mode when the application is opened, paused, or exited, without affecting the user's normal experience. Cover promotions are immersive full-screen promotions that appear below the application, displaying the promotion provider's avatar and name. Video stream promotions are a type of in-feed promotional information that uses video content to integrate products or services into the promotional information, such as displaying a product logo or product image in a corner of the video.

[0174] It should be noted that the second page image obtained during the promotional information testing process can also be used as a training sample for the promotional information type prediction model, thereby improving the model's prediction accuracy. For example, after capturing the content page to obtain the second page image, a second type label can be added to the second page image. Then, based on this second type label, the second page image can be processed for second feature annotation. This annotated second page image can then be used as a training sample for the promotional information type prediction model. Because the second page image obtained during the promotional information testing process can be used as a training sample for the promotional information type prediction model, the number and types of training samples for the model can be effectively expanded, which is beneficial for improving the model's prediction accuracy.

[0175] Step 123: When it is determined from the second prediction result that the second page image contains the first promotional information, the first promotional information is processed to obtain the promotional information judgment result.

[0176] In this step, since the second prediction result was obtained in step 122, it can be used to determine whether the second page image contains promotional information. If the second prediction result determines that the second page image contains first promotional information, since the first promotional information is not necessarily the target promotional information, information judgment processing can be performed on the first promotional information to obtain a promotional information judgment result. This allows subsequent steps to determine whether a promotional information test operation can be performed on the target promotional information based on the promotional information judgment result.

[0177] It should be noted that there are different implementation methods for processing the information of the first promotional information, and the appropriate method can be selected according to the actual application. This embodiment does not impose any specific limitations on this. For example, when the first promotional information is mainly text information, the text information in the first promotional information can be extracted first, and then the text information can be compared with the preset target text information for information judgment processing. As another example, when the first promotional information is mainly image information, the image information in the first promotional information can be identified first, and then the image information can be compared with the preset target image information for information judgment processing.

[0178] Reference Figure 8As shown in one embodiment of this application, the determination that the second page image contains the first promotional information based on the second prediction result in step 123 is further explained. When the second prediction result includes the prediction type and confidence level of the promotional information, the determination that the second page image contains the first promotional information based on the second prediction result in step 123 may include, but is not limited to, steps 1231 and 1232.

[0179] Step 1231: Compare the confidence level with the confidence threshold to obtain the second comparison result.

[0180] It should be noted that when the second prediction result includes the predicted type of promotional information and the confidence level, it means that the second page image may or may not contain promotional information corresponding to the target promotional information type. Whether the second page image contains promotional information corresponding to the target promotional information type needs to be determined based on the confidence level. Therefore, in this step, the confidence level can be compared with a confidence threshold to obtain a second comparison result, and then in subsequent steps, the second comparison result can be used to determine whether the second page image contains promotional information corresponding to the target promotional information type.

[0181] It should be noted that the confidence threshold is a preset threshold, which can be appropriately selected according to the actual application. This embodiment does not impose specific limitations on it. For example, the confidence threshold can be set to 0.5.

[0182] Step 1232: When the second comparison result is that the confidence level is greater than or equal to the confidence level threshold, it is determined that the second page image contains the first promotional information, and the promotional information type of the first promotional information is consistent with the promotional information prediction type.

[0183] In this step, since the second comparison result was obtained in step 1231, it can be used to determine whether the second page image contains promotional information corresponding to the target promotional information type. When the second comparison result has a confidence level greater than or equal to the confidence level threshold, it can be determined that the second page image contains the first promotional information, and the promotional information type of the first promotional information is consistent with the predicted promotional information type. Therefore, in subsequent steps, the first promotional information can be processed to obtain a promotional information judgment result, so that it can be used to determine whether to perform the promotional information test operation.

[0184] It should be noted that if the confidence level of the second comparison result is less than the confidence level threshold, it can be determined that the second page image does not contain promotional information, or the second page image does not contain promotional information corresponding to the prediction type of the promotional information. In this case, the promotional information testing operation can be exited to avoid performing invalid promotional information testing operations and improve the effectiveness of promotional information testing.

[0185] Reference Figure 9 As shown in one embodiment of this application, step 123 is further explained. When it is determined from the second prediction result that the second page image contains the first promotional information, and the test operation instruction includes a promotional information identifier, and the first promotional information includes promotional information text, step 123 may include, but is not limited to, steps 1233 to 1236.

[0186] Step 1233: Perform information recognition processing on the first promotional information to obtain the promotional information text;

[0187] Steps 1234: Send a promotional information identifier to the server so that the server can obtain the target text information based on the promotional information identifier;

[0188] Step 1235: Receive the target text information sent by the server;

[0189] Step 1236: Compare the promotional information text with the target text information. When the promotional information text matches the target text information, perform element information judgment processing on the first promotional information to obtain the promotional information judgment result.

[0190] It should be noted that although the second prediction result determines that the second page image contains the first promotional information, it is unclear whether the first promotional information is the target promotional information required for promotional information testing. In order to conduct promotional information testing more accurately, in this embodiment, the promotional information text in the first promotional information is first identified, and then the promotional information text is compared with the target text information of the target promotional information. When the two are consistent, it indicates that the first promotional information may be the target promotional information. Then, the element information judgment processing of the first promotional information can be performed to obtain the promotional information judgment result of whether the content page contains the target promotional information, so that subsequent steps can determine whether to perform the promotional information testing operation based on the promotional information judgment result. Specifically, since the test operation instruction includes a promotional information identifier, which is the identifier of the target promotional information required for the promotional information test, in order to determine whether the first promotional information is the target promotional information, in this embodiment, the first promotional information can first be processed by information recognition to obtain the promotional information text, and then the promotional information identifier can be sent to the server, so that the server can obtain the target text information based on the promotional information identifier. Then, the server sends the target text information. Since the target text information is the specific text information of the target promotional information, the promotional information text can be compared with the target text information. When the promotional information text matches the target text information, it indicates that the first promotional information is promotional information with the same text information as the target promotional information. Therefore, the first promotional information can be further processed by element information judgment to obtain the result of whether the content page contains the target promotional information. In addition, if the promotional information text does not match the target text information, it indicates that the first promotional information is not the target promotional information required for the promotional information test. At this time, the promotional information test operation can be exited to avoid performing invalid promotional information test operations and improve the effectiveness of the promotional information test.

[0191] It should be noted that there are different implementation methods for processing the first promotional information to obtain the promotional information text, and this embodiment does not impose a specific limitation on this method. For example, in one optional implementation, the interface elements in the first promotional information can be traversed to identify whether the interface elements in the first promotional information include a target interface element for displaying text information. When a target interface element is identified, the text information in the target interface element is extracted to obtain the promotional information text. Alternatively, in another optional implementation, the text information in the first promotional information can be identified using OCR technology to obtain the promotional information text.

[0192] It should be noted that the server can store different promotional information identifiers, different promotional information slots, and different promotional information text messages. The promotional information identifiers, promotional information slots, and promotional information text messages are all interconnected. For example, a given promotional information identifier corresponds to one promotional information slot and one promotional information text message. Therefore, when the server receives a promotional information identifier, it can retrieve the corresponding promotional information text message based on that identifier.

[0193] Reference Figure 10 As shown in the embodiment of this application, the step 1236, "performing element information judgment processing on the first promotion information to obtain the promotion information judgment result", is further explained. The step 1236, "performing element information judgment processing on the first promotion information to obtain the promotion information judgment result", may include, but is not limited to, steps 12361 and 12362.

[0194] Step 12361: Input the second page image into the element information detection model to obtain the third detection result;

[0195] Step 12362: Compare the third detection result with the preset target element information. When the third detection result is consistent with the target element information, the first promotion information is the target promotion information.

[0196] It should be noted that although comparing the promotional information text with the target text information yields a result indicating that the promotional information text and the target text information are consistent, this only means that the text information (such as copywriting content) in the first promotional information is the same as the text information (such as copywriting content) of the target promotional information required for promotional information testing. However, two promotional information pieces with the same text information may have different interface styles, and the coordinate positions of interface elements in different interface styles are also different. Therefore, two promotional information pieces with the same text information are not necessarily the same promotional information. Therefore, in order to determine whether the first promotional information is the target promotional information required for promotional information testing, and in order to conduct promotional information testing more accurately, in this embodiment, the second page image can be first input into the element information detection model to obtain a third detection result, and then the third detection result can be compared with the preset target element information. When the third detection result is consistent with the target element information, it indicates that the first promotional information is the target promotional information required for promotional information testing. Therefore, the promotional information judgment result that the first promotional information is the target promotional information can be obtained.

[0197] It should be noted that the element information detection model can be implemented using the YOLO algorithm. Its output third detection result can include element detection information of the interface elements of the first promotional information in the second page image, such as element coordinate detection information, element length detection information, and element width detection information. Among them, the element coordinate detection information can be the center coordinate detection information or the corner coordinate detection information of the element. Therefore, by comparing the element detection information with the preset target element information, it can be determined whether the element detection information is consistent with the preset target element information. When it is determined that the element detection information is consistent with the preset target element information, the promotional information judgment result that the first promotional information is the target promotional information can be obtained. In a specific example, suppose the first promotional information in the second page image includes interface element A and interface element B. After inputting the second page image into the element information detection model, the element coordinate detection information, element length detection information, and element width detection information of interface element A, as well as the element coordinate detection information, element length detection information, and element width detection information of interface element B, are obtained. Then, these detection information are compared with preset target element information, which includes the element coordinate information, element length information, and element width information of multiple interface elements. When all these detection information can find consistent content in the preset target element information, the first promotional information in the second page image can be considered as the target promotional information required for promotional information testing.

[0198] It should be noted that the second page image obtained during the promotional information testing process can also be used as a training sample for the element information detection model, thereby improving the model's detection accuracy. For example, after capturing the content page to obtain the second page image, a third-type label can be added to the image. Then, based on this third-type label, the image can undergo third-feature annotation processing. This annotated second page image can then be used as a training sample for the element information detection model. Because the second page image obtained during the promotional information testing process can be used as a training sample, the number and types of training samples for the element information detection model can be effectively expanded, which is beneficial for improving the model's detection accuracy.

[0199] Reference Figure 11 As shown in the embodiment of this application, step 130 is further described. Step 130 may include, but is not limited to, steps 131 to 133.

[0200] Step 131: Input the image of the second page into the trigger area prediction model to obtain the trigger area prediction information of the target promotion information;

[0201] Step 132: Determine the trigger area in the target promotion information based on the trigger area prediction information;

[0202] Step 133: Perform a trigger test based on the trigger area.

[0203] In this embodiment, after performing step 120 to determine the promotional information of the content page and obtaining the promotional information determination result, if the promotional information determination result indicates that the content page contains target promotional information, a trigger test can be performed on the target promotional information. However, since the target promotional information may include image information, text information, link information, avatar information, and nickname information, etc., text information generally cannot link to other pages, while image information, link information, avatar information, and nickname information can generally link to different pages. For example, image information and link information can link to the promotional information page, and avatar information and nickname information can link to the promotional information provider's profile page. Therefore, it is necessary to identify the linkable information in the target promotional information in order to accurately perform the promotional information test operation through this linkable information. Since interface elements are information carriers used to display promotional information, identifying the trigger areas in the target promotional information can effectively identify the linkable information in the target promotional information. Therefore, in this embodiment, the second page image can be input into the trigger area prediction model to obtain the trigger area prediction information of the target promotional information. Then, the trigger area in the target promotional information can be determined based on the trigger area prediction information. Finally, the trigger test can be performed based on the trigger area to achieve the purpose of accurately executing the promotional information test operation.

[0204] It should be noted that the trigger area prediction model can be implemented using the YOLO algorithm. The trigger area prediction information output by the model can include the predicted coordinates, length, and width of the trigger area. The predicted coordinates can be the center coordinates or the corner coordinates of the trigger area. Therefore, the trigger areas in the target promotion information can be determined based on these predicted trigger areas, so that trigger testing can be performed based on these trigger areas, thereby improving the effectiveness of the promotion information testing.

[0205] It should be noted that the second page image obtained during the promotional information testing process can also be used as a training sample for the trigger region prediction model, thereby improving its prediction accuracy. For example, after capturing the content page to obtain the second page image, a fourth type of label can be added to the second page image. Then, based on this fourth type of label, the second page image can be processed for fourth feature annotation. This annotated second page image can then be used as a training sample for the trigger region prediction model. Because the second page image obtained during the promotional information testing process can be used as a training sample for the trigger region prediction model, the number and types of training samples for the model can be effectively expanded, which is beneficial for improving its prediction accuracy.

[0206] Reference Figure 12 As shown in the embodiment of this application, step 110 is further described. When the test operation instruction includes a promotional information identifier, step 110 may include, but is not limited to, steps 111 to 114.

[0207] Step 111: In response to the test operation command, send a promotion information identifier to the server so that the server can obtain the target promotion information position information based on the promotion information identifier;

[0208] Step 112: Receive the target promotion information bit information sent by the server;

[0209] Step 113: Determine the target promotion information entry point based on the target promotion information position information;

[0210] Step 114: Enter the content page through the target promotion information entry.

[0211] In this embodiment, when the test operation instruction includes a promotional information identifier, the target promotional information entry point required for the promotional information test can be determined first based on the promotional information identifier. Then, the corresponding content page can be accessed through the target promotional information entry point, so that subsequent steps can execute the corresponding promotional information test operation. Since the server can store content such as the promotional information identifier, the promotional information position information corresponding to the promotional information identifier, and the promotional information text corresponding to the promotional information identifier, when a test operation instruction including a promotional information identifier is received from the operation terminal, the server can respond to the test operation instruction by sending the promotional information identifier to the server, so that the server can obtain the corresponding target promotional information position information based on the promotional information identifier, and then receive the target promotional information position information sent by the server. Next, the corresponding target promotional information entry point is determined based on the target promotional information position information, and then the corresponding content page is accessed through the target promotional information entry point, so that subsequent steps can execute the corresponding promotional information test operation based on the content page.

[0212] Furthermore, in one embodiment of this application, the promotional information testing method is further described. After performing step 160, the promotional information testing method may also include, but is not limited to, the following steps:

[0213] Re-execute the trigger test on the target promotion information;

[0214] After re-executing the trigger test, a new first page image is obtained by capturing the screen display content again;

[0215] The new first page image is input into the promotion information page detection model to obtain the new first detection result;

[0216] The correctness of the new first page image is determined based on the new first detection result until the number of trigger tests reaches a preset threshold.

[0217] In this embodiment, in order to achieve more accurate and effective promotional information testing, an upper limit (i.e., a preset threshold) can be preset for the number of times promotional information testing can be performed. After determining the correctness of the first page image based on the first detection result each time, the trigger test is re-executed on the target promotional information. After re-executing the trigger test, the screen display content is re-captured to obtain a new first page image. Then, the new first page image is input into the promotional information page detection model to obtain a new first detection result. Next, the correctness of the new first page image is determined based on the new first detection result, until the number of trigger tests reaches the preset threshold.

[0218] It should be noted that during the promotional information testing process, if a trigger test is performed on a specific trigger area of ​​the target promotional information, and the resulting first page image is then input into the promotional information page detection model to obtain a first detection result, the correctness of the first page image is determined based on this first detection result. In other words, after completing the promotional information testing operation for a specific trigger area, a trigger test can be performed on another trigger area of ​​the target promotional information. The resulting new first page image is then input into the promotional information page detection model to obtain a new first detection result. The correctness of the new first page image is then determined based on this new first detection result, and so on, until the number of trigger tests reaches a preset threshold. Furthermore, if the number of trigger areas in the target promotional information is less than the preset threshold, and all trigger areas of the target promotional information have completed the promotional information testing operation, the promotional information testing operation can be stopped. At this point, relevant screenshots and test results from the promotional information testing process can be uploaded to the server for storage. This allows subsequent steps to use these screenshots and test results for testing analysis, optimization of interactive applications, and other processing.

[0219] It should be noted that the preset threshold can be appropriately selected according to the actual application situation, and this embodiment does not impose specific limitations on it.

[0220] To more clearly illustrate the processing flow of the promotional information testing method provided in the embodiments of this application, specific examples are given below.

[0221] Example 1:

[0222] like Figure 13 As shown, Figure 13 This is a flowchart illustrating a specific example of a promotional information testing method provided in this application. Figure 13 The promotional information testing method includes steps 210 to 340.

[0223] Step 210: Enter the promotion information identifier in the operation terminal, and send a test operation instruction including the promotion information identifier to the test terminal.

[0224] Step 220: The test terminal obtains the promotion information entry information corresponding to the promotion information identifier from the server according to the promotion information identifier in the test operation instruction.

[0225] Step 230: The test terminal determines the target promotion information entry point based on the promotion information entry point information.

[0226] Step 240: The test terminal enters the corresponding content page through the target promotion information entry.

[0227] It should be noted that when the test terminal is an Android device, the Uiautomator automation framework can be used to trigger the test terminal to enter the corresponding content page through the target promotion information entry point; when the test terminal is an iOS device, the Webdriveragent automation framework can be used to trigger the test terminal to enter the corresponding content page through the target promotion information entry point.

[0228] Step 250: Capture the content page to obtain the second page image, and input the second page image into the promotion information type prediction model to obtain the second prediction result.

[0229] Step 260: Determine whether the second page image contains the first promotional information based on the second prediction result. If yes, proceed to step 270; otherwise, exit the promotional information test operation.

[0230] In this step, when it is determined from the second prediction result that the second page image contains, for example... Figure 14 The promotional information interface 104 shown indicates that the second page image contains the first promotional information, wherein... Figure 14 The promotion information interface 104 is used to display the first promotion information.

[0231] Step 270: Identify the promotional information text in the first promotional information, obtain the target text information from the server according to the promotional information identifier, compare the promotional information text with the target text information, if the comparison result is that the two are consistent, proceed to step 280, otherwise, exit the promotional information test operation.

[0232] Step 280: Input the second page image into the element information detection model to obtain the third detection result.

[0233] In this step, after the second page image is input into the element information detection model, the model outputs element detection information for the interface elements in the first promotional information. This element detection information may include element coordinate detection information, element length detection information, and element width detection information for different interface elements. Therefore, based on this element detection information, all interface elements in the first promotional information can be determined, for example... Figure 15 As shown, the first promotional information can be identified as including a first interface element 1041, a second interface element 1042, a third interface element 1043, a fourth interface element 1044, a fifth interface element 1045, and a sixth interface element 1046.

[0234] Step 290: Compare the third detection result with the preset target element information. If the comparison result is consistent, determine the first promotion information as the target promotion information and execute step 300. Otherwise, exit the promotion information test operation.

[0235] Step 300: Input the second page image into the trigger area prediction model to identify the trigger area in the target promotion information.

[0236] Step 310: Randomly select a trigger area from the target promotion information and perform a trigger test based on the selected trigger area.

[0237] Step 320: After executing the trigger test, capture the screen display content to obtain the first page image, and input the first page image into the promotion information page detection model to obtain the first detection result.

[0238] It should be noted that since different trigger areas in the target promotional information link to different pages, in this step, after performing the trigger test, it is necessary to capture a screenshot of the screen display to obtain an image of the first page. This is so that subsequent steps can determine whether the first page image is the correct promotional information page. For example... Figure 16 As shown, after performing a trigger test on the profile picture information in the target promotional information, the correct screen display content is: Figure 16 The promotional information provider's profile page.

[0239] Step 330: Compare the first detection result with the preset target promotion information page. If the comparison result is consistent, proceed to step 340; otherwise, exit the promotion information test operation.

[0240] Step 340: Return to the initial content page from the current screen display and jump to step 310 until the number of trigger tests reaches the preset threshold.

[0241] Example 2:

[0242] like Figure 17 As shown, Figure 17 This is a flowchart illustrating the promotional information testing method provided in another specific example of this application. Figure 17 The promotional information testing method includes the training process of the AI ​​model and the process of using the AI ​​model to conduct automated promotional information testing.

[0243] When the test terminal executes the AI ​​model training process, it captures relevant page images, including the second page image of the content page accessed through the target promotion information entry and the first page image of the screen display content after the trigger test is executed. Then, the test terminal performs feature annotation processing on the first page image and the second page image respectively, and then inputs the feature-annotated first page image and the second page image into the corresponding AI model for model training. For example, the feature-annotated first page image is input into the promotion information page detection model for model training, the feature-annotated second page image is input into the promotion information type prediction model for model training, the feature-annotated second page image is input into the element information detection model for model training, and the feature-annotated second page image is input into the trigger area prediction model for model training.

[0244] When a test terminal uses an AI model to test promotional information, the test terminal automates the process by integrating the AI ​​model. For example, suppose the test terminal sends a Hypertext Transfer Protocol (HTTP) request to the YOLO server via a test interface to request the recognition of the first page image. The YOLO server can be a functional component installed on the test terminal or on the server. Upon receiving the HTTP request, the YOLO server sends a prediction trigger message to the message queue module to subscribe to prediction results. The Keras neural network library, which supports the YOLO network model, reads the prediction trigger message from the message queue module and retrieves the corresponding prediction result from the relevant AI model (such as a promotional information type prediction model). The retrieved prediction result is then returned to the message queue module, which in turn returns it to the YOLO server. The YOLO server then returns the prediction result to the test terminal via the test interface, enabling the test terminal to perform promotional information testing based on the prediction result.

[0245] It should be noted that the YOLO server sends prediction trigger messages to the message queue module to subscribe to prediction results. The specific process is as follows: The YOLO server reads the AI ​​model (the AI ​​model implemented by the YOLO algorithm) corresponding to the model classification information carried in the HTTP request from the relevant configuration file. Then, it uses the model name of the AI ​​model as the title name of the prediction trigger message and sends the prediction trigger message to the message queue module. At the same time, the YOLO server listens for the prediction results corresponding to the title name. When the prediction result corresponding to the title name is detected, it means that the corresponding prediction processing has been completed. At this time, the YOLO server can read the prediction result corresponding to the title name from the message queue module.

[0246] It should be noted that the Keras Neural Networks library is an open-source artificial neural network library written in Python that provides high-level application programming interfaces for the design, debugging, evaluation, application, and visualization of deep learning models.

[0247] Example 3:

[0248] like Figure 18 As shown, Figure 18 This is a flowchart illustrating the promotional information testing method provided in another specific example of this application. Figure 18 The promotional information testing method includes the process of obtaining training samples for the AI ​​model and the process of using the AI ​​model to conduct automated promotional information testing.

[0249] When the test terminal acquires training samples for the AI ​​model, it captures relevant page images and adds them to the AI ​​model's training sample set for training. For example, the test terminal first obtains promotional information based on the promotional information identifier and displays this information on the content page accessed through the target promotional information entry point. Then, the test terminal searches for this promotional information on the content page by searching for the words "promotional information." Once found, the test terminal takes a screenshot of the content page to obtain the first training sample and adds it to the AI ​​model's training sample set for training. Additionally, after finding the promotional information on the content page, the test terminal determines the click location within the promotional information and performs a click operation at that location. This causes the content page to redirect. After the redirection, the test terminal captures the content displayed on the screen after the redirection to obtain the second training sample and adds it to the AI ​​model's training sample set for training.

[0250] When the test terminal uses an AI model to conduct automated promotional information testing, it first obtains the promotional information based on the promotional information identifier and displays it on the content page accessed through the target promotional information entry point. Then, the test terminal uses the AI ​​model to search for the promotional information on the content page. Once the promotional information is found and confirmed as the target promotional information, the test terminal uses the AI ​​model to determine the click location within the promotional information and executes a click operation at that location. At this point, the content page will redirect. After the content page redirects, the test terminal captures a screenshot of the displayed content. Next, the test terminal uses the AI ​​model to predict the screenshot and obtain a prediction result, then uses this prediction result to determine the correctness of the screenshot. Furthermore, during the automated promotional information testing process, the test terminal also uses the relevant screenshot content and the AI ​​model's prediction results as training samples, adding these training samples to the AI ​​model's training sample set for further training.

[0251] Reference Figure 19 This application also discloses a promotional information testing device, which can be used as such Figure 1 The test terminal 102 in the illustrated embodiment implements the promotional information testing method as described in the previous embodiment. The promotional information testing device 400 includes:

[0252] Interaction module 401 is used to respond to test operation commands and enter the content page through the target promotion information entry point;

[0253] The promotion information judgment module 402 is used to perform promotion information judgment processing on the content page and obtain the promotion information judgment result.

[0254] The trigger test module 403 is used to perform a trigger test on the target promotion information when the result of the promotion information judgment is that the content page contains the target promotion information. The trigger test is used to cause the content page to redirect.

[0255] The first capture module 404 is used to capture the screen display content to obtain the first page image after the trigger test is executed;

[0256] The page detection module 405 is used to input the first page image into the promotion information page detection model to obtain the first detection result of the page elements in the first page image, wherein the page elements are the information-carrying elements in the page used to display page information;

[0257] The result determination module 406 is used to determine the correctness of the first page image based on the first detection result and the page elements of the preset target promotion information page.

[0258] In one embodiment, the result determination module 406 includes:

[0259] The first comparison module is used to compare the first detection result with the page elements of the preset target promotion information page to obtain the first comparison result;

[0260] The correctness determination module is used to determine the correctness of the first page image based on the first comparison result.

[0261] In one embodiment, the first comparison module includes at least one of the following:

[0262] The first comparison submodule is used to compare the page element type detection information with the preset page element type information of the target promotion information page when the first detection result includes page element type detection information.

[0263] or,

[0264] The second comparison submodule is used to compare the page element layout detection information with the preset page element layout information of the target promotion information page when the first detection result includes page element layout detection information.

[0265] In one embodiment, the correctness determination module includes:

[0266] The first determination submodule is used to determine that the first page image is correct when the first comparison result is that the first detection result is consistent with the page elements of the target promotion information page;

[0267] or,

[0268] The second determination submodule is used to determine that the first page image is incorrect when the first comparison result is that the first detection result is inconsistent with the page elements of the target promotion information page.

[0269] In one embodiment, the promotion information judgment module 402 includes:

[0270] The second capture module is used to capture the content page to obtain the second page image;

[0271] The type prediction module is used to input the second page image into the promotional information type prediction model to obtain the second prediction result;

[0272] The information judgment module is used to perform information judgment processing on the first promotional information when it is determined from the second prediction result that the second page image contains the first promotional information, and to obtain the promotional information judgment result.

[0273] In one embodiment, when the test operation instruction includes a promotional information identifier and the first promotional information includes promotional information text, the information judgment module includes:

[0274] The information recognition module is used to perform information recognition processing on the first promotional information to obtain the promotional information text;

[0275] The first sending module is used to send a promotional information identifier to the server, so that the server can obtain the target text information based on the promotional information identifier.

[0276] The first receiving module is used to receive target text information sent by the server;

[0277] The information judgment submodule is used to compare the promotional information text with the target text information. When the promotional information text matches the target text information, the element information judgment processing is performed on the first promotional information to obtain the promotional information judgment result.

[0278] In one embodiment, the information determination submodule includes:

[0279] The element detection module is used to input the second page image into the element information detection model to obtain the third detection result;

[0280] The second comparison module is used to compare the third detection result with the preset target element information. When the third detection result is consistent with the target element information, the first promotion information is determined to be the target promotion information.

[0281] In one embodiment, when the second prediction result includes the prediction type and confidence level of the promotion information, the information judgment module includes:

[0282] The third comparison module is used to compare the confidence level with the confidence threshold to obtain the second comparison result;

[0283] The promotion information determination module is used to determine that the second page image contains first promotion information when the second comparison result is a confidence level greater than or equal to a confidence level threshold, and the promotion information type of the first promotion information is consistent with the promotion information prediction type.

[0284] In one embodiment, the trigger test module 403 includes:

[0285] The region prediction module is used to input the second page image into the trigger region prediction model to obtain the trigger region prediction information of the target promotion information;

[0286] The region determination module is used to determine the trigger region in the target promotion information based on the trigger region prediction information;

[0287] The Trigger Test submodule is used to execute trigger tests based on the trigger area.

[0288] In one embodiment, when the test operation instruction includes a promotional information identifier, the interaction module 401 includes:

[0289] The second sending module is used to send a promotion information identifier to the server in response to a test operation command, so that the server can obtain the target promotion information position information based on the promotion information identifier.

[0290] The second receiving module is used to receive the target promotion information bit information sent by the server;

[0291] The entry point determination module is used to determine the entry point for the target promotion information position based on the target promotion information position information;

[0292] The page entry module is used to access the content page through the target promotion information entry point.

[0293] In one embodiment, the trigger test module 403 is further configured to re-execute the trigger test on the target promotion information;

[0294] The first capture module 404 is also used to capture the screen display content again to obtain a new first page image after re-executing the trigger test;

[0295] The page detection module 405 is also used to input the new first page image into the promotion information page detection model to obtain a new first detection result;

[0296] The result determination module 406 is also used to determine the correctness of the new first page image based on the new first detection result, until the number of trigger tests reaches a preset threshold.

[0297] In one embodiment, the promotion information testing device 400 further includes:

[0298] The first tag adding module is used to add the first type of tag to the image on the first page;

[0299] The first feature annotation module is used to perform first feature annotation processing on the first page image according to the first type of label;

[0300] The first model training module is used to train the promotion information page detection model by using the first page image, which has undergone the first feature annotation process, as a training sample.

[0301] In one embodiment, the promotion information testing device 400 further includes:

[0302] The second tag adding module is used to add a second type of tag to the images on the second page;

[0303] The second feature annotation module is used to perform second feature annotation processing on the second page image based on the second type of label;

[0304] The second model training module is used to train the promotion information type prediction model by using the second page image, which has undergone second feature annotation processing, as a training sample.

[0305] In one embodiment, the promotion information testing device 400 further includes:

[0306] The third tag adding module is used to add a third type of tag to the images on the second page;

[0307] The third feature annotation module is used to perform third feature annotation processing on the second page image based on the third type of label;

[0308] The third model training module is used to train the element information detection model by using the second page image, which has undergone third feature annotation processing, as a training sample.

[0309] In one embodiment, the promotion information testing device 400 further includes:

[0310] The fourth tag adding module is used to add a fourth type of tag to the images on the second page;

[0311] The fourth feature annotation module is used to perform fourth feature annotation processing on the second page image based on the fourth type of label;

[0312] The fourth model training module is used to train the trigger region prediction model by using the second page image, which has undergone fourth feature annotation processing, as a training sample.

[0313] It should be noted that since the promotion information testing device 400 in this embodiment can implement the promotion information testing method with the test terminal as the execution subject in the previous embodiment, the promotion information testing device 400 in this embodiment has the same technical principle and the same beneficial effect as the promotion information testing method with the test terminal as the execution subject in the previous embodiment. To avoid repetition, it will not be described again here.

[0314] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0315] Reference Figure 20 This application also discloses a promotional information testing device 500, which includes:

[0316] At least one processor 501;

[0317] At least one memory 502 is used to store at least one program;

[0318] When at least one program is executed by at least one processor 501, the promotional information testing method as described in any of the preceding embodiments is implemented.

[0319] This application also discloses a computer-readable storage medium storing a processor-executable program, which, when executed by a processor, is used to implement the promotional information testing method as described in any of the preceding embodiments.

[0320] This application also discloses a computer program product, including a computer program or computer instructions, which are stored in a computer-readable storage medium. The processor of a computer device reads the computer program or computer instructions from the computer-readable storage medium and executes the computer program or computer instructions, causing the computer device to perform the promotional information testing method as described in any of the preceding embodiments.

[0321] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatuses.

[0322] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0323] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, apparatuses, or units, and may be electrical, mechanical, or other forms.

[0324] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0325] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0326] If the integrated unit is implemented as 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 this application, in essence, or the part that contributes to the prior art, or all or part of the 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 to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0327] The step numbers in the above method embodiments are set only for ease of explanation and do not limit the order of the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

Claims

1. A method for testing promotional information, characterized in that, Includes the following steps: In response to a test operation command, the user enters the content page through the target promotional information entry point, wherein the test operation command includes a promotional information identifier. The second page image is obtained by cropping the content page; The second page image is input into the promotional information type prediction model to obtain the second prediction result; When it is determined from the second prediction result that the second page image contains first promotional information, the first promotional information is processed by information recognition to obtain the promotional information text; The system sends the promotional information identifier to the server, enabling the server to obtain target text information based on the promotional information identifier; receives the target text information sent by the server; compares the promotional information text with the target text information; if the promotional information text matches the target text information, inputs the second page image into the element information detection model to obtain a third detection result; compares the third detection result with preset target element information; if the third detection result matches the target element information, obtains a promotional information judgment result that the first promotional information is the target promotional information. When the determination result of the promotion information is that the content page contains target promotion information, a trigger test is performed on the target promotion information, wherein the trigger test is used to cause the content page to redirect; After the trigger test is executed, the screen display content is captured to obtain the first page image; The first page image is input into the promotion information page detection model to obtain the first detection result of the page elements in the first page image, wherein the page elements are information-carrying elements in the page used to display page information; The correctness of the first page image is determined based on the first detection result and the page elements of the preset target promotion information page.

2. The promotional information testing method according to claim 1, characterized in that, The step of determining the correctness of the first page image based on the first detection result and the page elements of the preset target promotion information page includes: The first detection result is compared with the page elements of the preset target promotion information page to obtain the first comparison result; The correctness of the first page image is determined based on the first comparison result.

3. The promotional information testing method according to claim 2, characterized in that, The first detection result is compared with the page elements of the preset target promotion information page. Includes at least one of the following steps: When the first detection result includes page element type detection information, the page element type detection information is compared with the preset page element type information of the target promotion information page; or, When the first detection result includes page element layout detection information, the page element layout detection information is compared with the preset page element layout information of the target promotion information page.

4. The promotional information testing method according to claim 2, characterized in that, Determining the correctness of the first page image based on the first comparison result includes: When the first comparison result is that the first detection result is consistent with the page elements of the target promotion information page, the first page image is determined to be correct; or, If the first comparison result shows that the first detection result is inconsistent with the page elements of the target promotion information page, then the first page image is determined to be incorrect.

5. The promotional information testing method according to claim 1, characterized in that, The second prediction result includes the type of promotional information prediction and the confidence level; The step of determining that the second page image contains the first promotional information based on the second prediction result includes: The confidence level is compared with the confidence threshold to obtain a second comparison result; When the second comparison result is that the confidence level is greater than or equal to the confidence level threshold, it is determined that the second page image contains first promotional information, and the promotional information type of the first promotional information is consistent with the promotional information prediction type.

6. The promotional information testing method according to claim 1, characterized in that, The trigger test performed on the target promotion information includes: The second page image is input into the trigger area prediction model to obtain the trigger area prediction information of the target promotion information; The triggering region in the target promotion information is determined based on the triggering region prediction information; Perform a trigger test based on the trigger area.

7. The promotional information testing method according to claim 1, characterized in that, The test operation instructions include promotional information identifiers; The response to the test operation command, entering the content page through the target promotion information entry point, includes: In response to a test operation command, the promotion information identifier is sent to the server, enabling the server to obtain the target promotion information bit information based on the promotion information identifier; Receive the target promotion information bit information sent by the server; The target promotion information entry point is determined based on the target promotion information position information; Access the content page through the target promotional information entry point.

8. The promotional information testing method according to claim 1, characterized in that, After determining the correctness of the first page image based on the first detection result, the promotional information testing method further includes: Re-execute the trigger test on the target promotion information; After re-executing the trigger test, a new first page image is obtained by capturing the screen display content again; The new first page image is input into the promotion information page detection model to obtain a new first detection result; The correctness of the new first page image is determined based on the new first detection result until the number of trigger tests reaches a preset threshold.

9. The promotional information testing method according to claim 1, characterized in that, The promotional information testing method also includes: Add a first type of tag to the image on the first page; The first page image is subjected to first feature annotation processing based on the first type of label; The first page image, which has undergone the first feature annotation process, is used as a training sample for the promotional information page detection model to train the model.

10. The promotional information testing method according to claim 1, characterized in that, The promotional information testing method also includes: Add a second type of tag to the image on the second page; The second page image is processed with second feature annotation based on the second type of label; The second page image, which has undergone the second feature annotation process, is used as a training sample for the promotion information type prediction model to train the model.

11. The promotional information testing method according to claim 1, characterized in that, The promotional information testing method also includes: Add a third type of tag to the image on the second page; The second page image is processed with third feature annotation based on the third type of label; The second page image, which has undergone the third feature annotation process, is used as a training sample for the element information detection model to train the model.

12. The promotional information testing method according to claim 6, characterized in that, The promotional information testing method also includes: Add a fourth type of tag to the image on the second page; The second page image is processed with fourth feature annotation based on the fourth type of label; The second page image, which has undergone the fourth feature annotation process, is used as a training sample for the trigger region prediction model, and the trigger region prediction model is trained accordingly.

13. A promotional information testing device, characterized in that, include: The interaction module is used to respond to test operation instructions and enter the content page through the target promotion information entry point. The test operation instructions include the promotion information identifier. The promotion information judgment module is used to perform promotion information judgment processing on the content page and obtain the promotion information judgment result; The promotion information judgment module includes: The second capture module is used to capture the content page to obtain a second page image; The type prediction module is used to input the second page image into the promotion information type prediction model to obtain a second prediction result; The information judgment module is used to: when it is determined from the second prediction result that the second page image contains first promotional information, perform information recognition processing on the first promotional information to obtain the promotional information text; send the promotional information identifier to the server, so that the server obtains target text information based on the promotional information identifier; receive the target text information sent by the server; compare the promotional information text with the target text information; when the promotional information text matches the target text information, input the second page image into the element information detection model to obtain a third detection result; compare the third detection result with preset target element information; when the third detection result matches the target element information, obtain a promotional information judgment result that the first promotional information is the target promotional information. The trigger test module is used to perform a trigger test on the target promotion information when the promotion information determination result is that the content page contains target promotion information, wherein the trigger test is used to cause the content page to redirect; The first capture module is used to capture the screen display content to obtain the first page image after the trigger test is executed; The page detection module is used to input the first page image into the promotion information page detection model to obtain the first detection result of the page elements in the first page image, wherein the page elements are information-carrying elements in the page used to display page information; The result determination module is used to determine the correctness of the first page image based on the first detection result and the page elements of the preset target promotion information page.

14. A promotional information testing device, characterized in that, include: At least one processor; At least one memory for storing at least one program; The promotional information testing method as described in any one of claims 1 to 12 is implemented when at least one of the programs is executed by at least one of the processors.

15. A computer-readable storage medium, characterized in that, It contains a processor-executable program, which, when executed by the processor, is used to implement the promotional information testing method as described in any one of claims 1 to 12.

16. A computer program product, comprising a computer program or computer instructions, characterized in that, The computer program or the computer instructions are stored in a computer-readable storage medium, the processor of the computer device reads the computer program or the computer instructions from the computer-readable storage medium, and the processor executes the computer program or the computer instructions, causing the computer device to perform the promotional information testing method as described in any one of claims 1 to 12.

Images