Game interface testing method and system

Abstract

The invention discloses a game interface test method and system, and the method comprises the steps: extracting normal data logic from a business layer interface, generating normal logic data according to the normal data logic, determining special data logic according to the normal data logic, and generating special logic data based on the special data logic through a deep learning model, and inputting the normal logic data and the special logic data into the business layer interface to obtain respective corresponding output data, and if the data representation forms of the output data of the normal logic data and the special logic data are consistent, indicating that the business layer interface can process the data according to the preset rule, so that the business layer interface is determined to be normal. The deep learning model is used for performing thinking divergence on the data conforming to the business layer interface logic to generate special logic data which comprises normal logic data, and the normal logic data and the special logic data are used for testing the business layer interface, so that the divergence of the data is improved, and the coverage rate of game interface testing is improved; therefore, the efficiency and reliability of game interface testing are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of interface testing, and particularly to a game interface testing method and system. Background Art

[0002] Interface testing is mainly used to detect the interaction points between external systems and systems as well as between internal subsystems, facilitating the discovery of underlying interface BUGs. At the same time, it can detect security and stability, and game interface testing is the key to ensuring the normal operation of game logic and the stability of user experience. In the prior art, for game interface testing, a large number of tests need to be carried out manually, with low efficiency. Moreover, due to the limitations of testers, data may be omitted, resulting in low test coverage and decreased reliability. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a game interface testing method and system that can improve the efficiency and reliability of game interface testing.

[0004] To solve the above technical problem, a technical solution adopted by the present invention is: A game interface testing method, comprising the steps of: Extracting normal data logic from the business layer interface and generating normal logic data according to the normal data logic; Determining special data logic according to the normal data logic and generating special logic data based on the special data logic using a deep learning model; Inputting the normal logic data and the special logic data into the business layer interface respectively to obtain first output data corresponding to the normal logic data and second output data corresponding to the special logic data; Judging whether the data presentation forms of the first output data and the second output data are the same. If so, it is determined that the business layer interface is normal; if not, it is determined that the business layer interface is abnormal.

[0005] To solve the above technical problem, another technical solution adopted by the present invention is: A game interface testing system, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented: Extracting normal data logic from the business layer interface and generating normal logic data according to the normal data logic; Determining special data logic according to the normal data logic and generating special logic data based on the special data logic using a deep learning model; Input the normal logic data and the special logic data into the service layer interface respectively, to obtain the first output data corresponding to the normal logic data and the second output data corresponding to the special logic data; Determine whether the data presentation forms of the first output data and the second output data are the same. If so, determine that the service layer interface is normal; if not, determine that the service layer interface is abnormal.

[0006] The beneficial effects of the present invention are as follows: Extract the normal data logic from the service layer interface, generate normal logic data according to the normal data logic, determine the special data logic according to the normal data logic, and use a deep learning model to generate special logic data based on the special data logic. Input the normal logic data and the special logic data into the service layer interface respectively, to obtain the first output data corresponding to the normal logic data and the second output data corresponding to the special logic data. If the data presentation forms of the first output data and the second output data are the same, it indicates that the service layer interface can process data according to the preset rules. Therefore, determine that the service layer interface is normal. In this way, use the deep learning model to perform divergent thinking on the data that conforms to the service layer interface logic, generate special logic data, and the special logic data contains the normal logic data. Use the normal logic data and the special logic data to test the service layer interface, improve the divergence of the data, avoid data omission, improve the coverage rate of the game interface test, and achieve better test results with less test time, thereby improving the efficiency and reliability of the game interface test. Description of the Drawings

[0007] Figure 1 It is a step flow chart of a game interface test method according to an embodiment of the present invention; Figure 2 It is a structural schematic diagram of a game interface test system according to an embodiment of the present invention; Figure 3 It is a schematic diagram of the splitting of the service layer interface in the game interface test method according to an embodiment of the present invention; Figure 4 It is a test schematic diagram of the game interface test method according to an embodiment of the present invention; Figure 5 It is a test schematic diagram of the logic layer interface in the game interface test method according to an embodiment of the present invention; Figure 6 It is a test schematic diagram of the data layer interface in the game interface test method according to an embodiment of the present invention. Detailed Embodiments

[0008] To describe the technical content, the achieved objectives and the effects of the present invention in detail, the following is described in conjunction with the embodiments and with reference to the drawings.

[0009] Please refer toFigure 1 , a game interface testing method, comprising the steps of: Extract normal data logic from the business layer interface and generate normal logic data according to the normal data logic; Determine special data logic according to the normal data logic, and use a deep learning model to generate special logic data based on the special data logic; Input the normal logic data and the special logic data into the business layer interface respectively, to obtain first output data corresponding to the normal logic data and second output data corresponding to the special logic data; Judge whether the data presentation forms of the first output data and the second output data are consistent. If so, determine that the business layer interface is normal; if not, determine that the business layer interface is abnormal.

[0010] As can be seen from the above description, the beneficial effects of the present invention are as follows: Extract normal data logic from the business layer interface, generate normal logic data according to the normal data logic, determine special data logic according to the normal data logic, and use a deep learning model to generate special logic data based on the special data logic. Input the normal logic data and the special logic data into the business layer interface respectively, to obtain first output data corresponding to the normal logic data and second output data corresponding to the special logic data. If the data presentation forms of the first output data and the second output data are consistent, it indicates that the business layer interface can process data according to the preset rules. Therefore, it is determined that the business layer interface is normal. In this way, the deep learning model is used to conduct divergent thinking on the data that conforms to the business layer interface logic, generate special logic data, and the special logic data includes the normal logic data. The normal logic data and the special logic data are used to test the business layer interface, improving the divergence of the data, avoiding data omission, increasing the coverage rate of the game interface test, achieving better test results with less test time, and thus improving the efficiency and reliability of the game interface test.

[0011] Further, the step of extracting normal data logic from the business layer interface and generating normal logic data according to the normal data logic includes: Split the business layer interface into a logic layer interface and a data layer interface; Extract first normal data logic from the logic layer interface and extract second normal data logic from the data layer interface; Generate first normal logic data according to the first normal data logic and generate second normal logic data according to the second normal data logic.

[0012] As can be seen from the above description, splitting the business layer interface into a logic layer interface and a data layer interface, where the logic layer interface is a data interface containing logic events split from the business layer, and the data layer interface mainly processes data storage, processing, and transmission. Extracting the corresponding normal data logics from the logic layer interface and the data layer interface respectively, and then generating their respective normal logic data according to their respective normal data logics can generate data that more precisely meets their respective requirements, avoiding test omissions or misjudgments caused by logic or data problems.

[0013] Further, the determining the special data logic according to the normal data logic and using a deep learning model to generate special logic data based on the special data logic includes: Determining a first special data logic according to the first normal data logic and using a deep learning model to generate a first special logic data based on the first special data logic; Determining a second special data logic according to the second normal data logic and using a deep learning model to generate a second special logic data based on the second special data logic.

[0014] As can be seen from the above description, using a deep learning model to generate special logic data corresponding to the logic layer interface and special logic data corresponding to the data layer interface respectively can more comprehensively test the robustness and stability of the game, discover potential vulnerabilities, anomalies, or undefined behaviors, thereby improving the reliability of the test.

[0015] Further, the respectively inputting the normal logic data and the special logic data into the business layer interface to obtain a first output data corresponding to the normal logic data and a second output data corresponding to the special logic data includes: Inputting the first normal logic data and the first special logic data into the logic layer interface respectively to obtain a first logic layer output data corresponding to the first normal logic data and a second logic layer output data corresponding to the first special logic data; Inputting the second normal logic data and the second special logic data into the data layer interface respectively to obtain a first data layer output data corresponding to the second normal logic data and a second data layer output data corresponding to the second special logic data.

[0016] As can be seen from the above description, inputting the normal logic data and the special logic output corresponding to the logic layer interface into the logic layer interface respectively, and inputting the normal logic data and the special logic output corresponding to the data layer interface into the data layer interface respectively to obtain their respective corresponding output data facilitates accurately and quickly locating and fixing problems subsequently.

[0017] Further, the determination of whether the data representation forms of the first output data and the second output data are consistent, and if so, determining that the service layer interface is normal, and if not, determining that the service layer interface is abnormal includes: Determine whether the data representation forms of the first logical layer output data and the second logical layer output data are consistent. If so, determine that the logical layer interface is normal. If not, determine that the logical layer interface is abnormal; Determine whether the data representation forms of the first data layer output data and the second data layer output data are consistent. If so, determine that the data layer interface is normal. If not, determine that the data layer interface is abnormal.

[0018] As can be seen from the above description, by comparing the data representation forms of the first logical layer output data and the second logical layer output data output by the logical layer interface, and then comparing the data representation forms of the first data layer output data and the second data layer output data output by the data layer interface, the interface consistency can be accurately detected. By comparing the representation forms of the output data, potential logical errors or data processing problems can be discovered, improving the effectiveness of game interface testing.

[0019] Please refer to Figure 2 , a game interface testing system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented: Extract the normal data logic from the service layer interface and generate normal logic data according to the normal data logic; Determine the special data logic according to the normal data logic, and use a deep learning model to generate special logic data based on the special data logic; Input the normal logic data and the special logic data into the service layer interface respectively to obtain the first output data corresponding to the normal logic data and the second output data corresponding to the special logic data; Determine whether the data representation forms of the first output data and the second output data are consistent. If so, determine that the service layer interface is normal. If not, determine that the service layer interface is abnormal.

[0020] As can be seen from the above description, the beneficial effects of the present invention are as follows: Extract the normal data logic from the business layer interface, generate normal logic data according to the normal data logic, determine the special data logic based on the normal data logic, and use a deep learning model to generate special logic data based on the special data logic. Input the normal logic data and the special logic data into the business layer interface respectively to obtain the first output data corresponding to the normal logic data and the second output data corresponding to the special logic data. If the data presentation forms of the first output data and the second output data are the same, it indicates that the business layer interface can process data according to the preset rules. Therefore, it is determined that the business layer interface is normal. Thus, the deep learning model is used to conduct divergent thinking on the data that conforms to the business layer interface logic to generate special logic data. The special logic data includes the normal logic data. The normal logic data and the special logic data are used to test the business layer interface, which improves the divergence of the data, avoids data omission, improves the coverage rate of the game interface test, and can achieve better test results with less test time, thereby improving the efficiency and reliability of the game interface test.

[0021] Further, the step of extracting the normal data logic from the business layer interface and generating normal logic data according to the normal data logic includes: Split the business layer interface into a logic layer interface and a data layer interface; Extract the first normal data logic from the logic layer interface and extract the second normal data logic from the data layer interface; Generate the first normal logic data according to the first normal data logic and generate the second normal logic data according to the second normal data logic.

[0022] As can be seen from the above description, splitting the business layer interface into a logic layer interface and a data layer interface, where the logic layer interface is a data interface obtained by splitting the business layer and containing logic events, and the data layer interface mainly processes data storage, processing, and transmission. Extract the corresponding normal data logic from the logic layer interface and the data layer interface respectively, and then generate the respective normal logic data according to their respective normal data logics, which can generate data that more precisely meets their respective requirements and avoid test omissions or misjudgments caused by logic or data problems.

[0023] Further, the step of determining the special data logic based on the normal data logic and using a deep learning model to generate special logic data based on the special data logic includes: Determine the first special data logic based on the first normal data logic and use a deep learning model to generate the first special logic data based on the first special data logic; Determine the second special data logic based on the second normal data logic and use a deep learning model to generate the second special logic data based on the second special data logic.

[0024] As can be seen from the above description, by separately using the deep learning model to generate the special logical data corresponding to the logical layer interface and the special logical data corresponding to the data layer interface, the robustness and stability of the game can be more comprehensively tested, potential vulnerabilities, anomalies or undefined behaviors can be discovered, thereby improving the reliability of the test.

[0025] Further, the step of respectively inputting the normal logical data and the special logical data into the service layer interface to obtain the first output data corresponding to the normal logical data and the second output data corresponding to the special logical data includes: Respectively input the first normal logical data and the first special logical data into the logical layer interface to obtain the first logical layer output data corresponding to the first normal logical data and the second logical layer output data corresponding to the first special logical data; Respectively input the second normal logical data and the second special logical data into the data layer interface to obtain the first data layer output data corresponding to the second normal logical data and the second data layer output data corresponding to the second special logical data.

[0026] As can be seen from the above description, by respectively inputting the normal logical data and the special logical output corresponding to the logical layer interface into the logical layer interface, and respectively inputting the normal logical data and the special logical output corresponding to the data layer interface into the data layer interface to obtain the corresponding output data respectively, it is convenient to accurately and quickly locate and fix problems subsequently.

[0027] Further, the step of determining whether the data presentation forms of the first output data and the second output data are consistent, if so, determining that the service layer interface is normal, and if not, determining that the service layer interface is abnormal includes: Determine whether the data presentation forms of the first logical layer output data and the second logical layer output data are consistent. If so, determine that the logical layer interface is normal. If not, determine that the logical layer interface is abnormal; Determine whether the data presentation forms of the first data layer output data and the second data layer output data are consistent. If so, determine that the data layer interface is normal. If not, determine that the data layer interface is abnormal.

[0028] As described above, by comparing the data presentation forms of the first logical layer output data and the second logical layer output data output by the logical layer interface, and then comparing the data presentation forms of the first data layer output data and the second data layer output data output by the data layer interface, the interface consistency can be accurately detected. By comparing the presentation forms of the output data, potential logical errors or data processing problems can be discovered, improving the effectiveness of game interface testing.

[0029] The above game interface testing method and system of the present invention can be applied to game interface testing scenarios, which will be described below through specific embodiments: Please refer to Figure 1 , Figures 3 - 6 , the first embodiment of the present invention is: A game interface testing method, including the steps: S1. Extract the normal data logic from the service layer interface and generate normal logic data according to the normal data logic, specifically including S11 - S13: S11. Split the service layer interface into a logical layer interface (denoted as LL) and a data layer interface (denoted as DL). As Figure 3 shown, the service layer interface is composed of a logical layer interface and a data layer interface, and is stored by a module group.

[0030] Among them, the service layer interface in the game interface contains the data involved in the game business. For example, the player purchase item interface in the client will involve specific NPCs (Non - Player Characters), specific maps, and specific scenes. The logical layer interface is a data interface containing logical events split from the service layer. For example, the interface for interaction within the client is the client interface, and the interface for interaction between the client and the server is the network communication interaction interface. For example, when a player goes to a map to buy something and can only buy on that map, the purchase behavior requires a relevance judgment on the involved logic, that is, data that conforms to the current business logic situation. The data layer interface is in line with the data presentation situation. For example, when a player purchases an item and an item is returned, this correct presentation meets the expectations and satisfies the needs of the data layer.

[0031] S12. Extract the first normal data logic from the logical layer interface, as Figure 5 shown, and extract the second normal data logic from the data layer interface, as Figure 6 shown.

[0032] Among them, the first normal data logic is the normal logic required by the logical layer interface, and the second normal data logic is the normal logic required by the data layer interface.

[0033] S13. Generate the first normal logical data according to the first normal data logic, and generate the second normal logical data according to the second normal data logic, as Figure 4 shown.

[0034] S2. Determine the special data logic according to the normal data logic, and use a deep learning model to generate special logical data based on the special data logic, specifically including S21 - S22: S21. Determine the first special data logic according to the first normal data logic, as Figure 5 shown, and use a deep learning model to generate the first special logical data based on the first special data logic.

[0035] S22. Determine the second special data logic according to the second normal data logic, and use a deep learning model to generate the second special logical data based on the second special data logic, as Figure 6 shown.

[0036] In an alternative embodiment, as Figure 4 shown, the deep learning model is a GAN (Generative Adversarial Networks) model.

[0037] For example, "0: false, 1: true" is the normal data logic, that is, 0 and 1 are used as the current logical basis when transmitting data. The special data logic is a data template generated by using GAN for thinking divergence based on the normal data logic, and is used to test the filtering and processing capabilities of the interface for data. The special data logic contains the normal data logic.

[0038] S3. Input the normal logical data and the special logical data into the service layer interface respectively, and obtain the first output data corresponding to the normal logical data and the second output data corresponding to the special logical data, specifically including S31 - S32: S31. Input the first normal logical data and the first special logical data into the logic layer interface respectively, and obtain the first logic layer output data corresponding to the first normal logical data and the second logic layer output data corresponding to the first special logical data.

[0039] S32. Input the second normal logical data and the second special logical data into the data layer interface respectively, and obtain the first data layer output data corresponding to the second normal logical data and the second data layer output data corresponding to the second special logical data.

[0040] S4. Determine whether the data presentation forms of the first output data and the second output data are the same. If so, determine that the service layer interface is normal; if not, determine that the service layer interface is abnormal. As Figure 4 shown, it specifically includes S41 - S42: S41. Determine whether the data presentation forms of the first output data of the logic layer and the second output data of the logic layer are the same. If so, determine that the logic layer interface is normal; if not, determine that the logic layer interface is abnormal.

[0041] S42. Determine whether the data presentation forms of the first output data of the data layer and the second output data of the data layer are the same. If so, determine that the data layer interface is normal; if not, determine that the data layer interface is abnormal.

[0042] When the game communicates with the server, it is stipulated that text is used for communication and decimals are not allowed. At this time, data types such as "0.1" and "x.x" are regarded as illegal data types. Under the divergent thinking of the deep learning model, it thinks of scientific notation, that is, using E for data expression. For example, 10E0 = 10. Under special data logic, it is found that the logic layer interface will process scientific notation, that is, 10E0 outputs 10, 10E - 1 outputs 1, and 10E - 2 outputs 0.1, which conforms to the special data logic. Then when the data passes through the logic layer interface, 0.1 is output, that is, the data defense of the logic layer interface is breached. After modification, the logic layer interface judges in an integer manner, taking data where value > 0.

[0043] That is to say, when the output data of the special data logic generated by the deep learning model is inconsistent with the output data presentation form of the normal data logic, it is regarded as a data presentation error. For example, it is known that the data packet data structure is: the first few data are the data size to be sent, and the following are the contents. For example, 01 00. Then the data type generated by the deep learning model is the packet body size + content. At this time, the normal data logic is: the specified packet body size + content, and the special data logic is: the enlarged packet body size + content; for example, the normal logical data 05 00 05 F0 03 under the normal data logic is a purchase data packet, where 05 F0 represents the merchant ID and 03 represents the quantity. The corresponding special logical data under the special data logic is 06 00 05 F0 03 00, which is the same as the purchase of 05 00 05 F0 03, but 06 00 05 F0 03 FF can be executed in the interface, that is, the quantity breaks through the regulation.

[0044] The above game interface testing method of the present invention uses a deep learning model to generate special logic data, combines the normal logic data to test the interface, increases the divergence of data, saves labor costs, and more comprehensively investigates potential game data security hazards, effectively improving the efficiency and reliability of game interface testing.

[0045] Please refer to Figure 2 , the second embodiment of the present invention is: A game interface testing system, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements each step in the game interface testing method in the first embodiment.

[0046] In summary, the present invention provides a game interface testing method and system. It extracts the normal data logic from the business layer interface, generates normal logic data according to the normal data logic, determines the special data logic based on the normal data logic, and uses a deep learning model to generate special logic data based on the special data logic. The normal logic data and the special logic data are respectively input into the business layer interface to obtain the first output data corresponding to the normal logic data and the second output data corresponding to the special logic data. If the data presentation forms of the first output data and the second output data are the same, it indicates that the business layer interface can process data according to the preset rules, so it is determined that the business layer interface is normal. Thus, the deep learning model is used to conduct divergent thinking on the data that conforms to the business layer interface logic to generate special logic data. The special logic data contains the normal logic data. The normal logic data and the special logic data are used to test the business layer interface, improving the divergence of data, avoiding data omission, increasing the coverage rate of game interface testing, achieving a better testing effect with less testing time, and thus improving the efficiency and reliability of game interface testing. In addition, the business layer interface is split into a logic layer interface and a data layer interface. The logic layer interface is a data interface split from the business layer that contains logical events. The data layer interface mainly processes data storage, processing, and transmission. The corresponding normal data logics are respectively extracted from the logic layer interface and the data layer interface, and then the respective normal logic data are generated according to their respective normal data logics, which can more accurately generate data that meets their respective requirements, avoiding testing omissions or misjudgments caused by logic or data problems. Moreover, using the deep learning model to generate the special logic data corresponding to the logic layer interface and the special logic data corresponding to the data layer interface respectively can more comprehensively test the robustness and stability of the game, and discover potential vulnerabilities, anomalies, or undefined behaviors.

[0047] The above are only the embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent transformation made using the specification and drawings of the present invention, or directly or indirectly applied in the relevant technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. A game interface testing method, characterized in that, Including the steps: Extract the normal data logic from the business layer interface and generate normal logic data according to the normal data logic; Determine the special data logic according to the normal data logic, and use a deep learning model to generate special logic data based on the special data logic; Input the normal logic data and the special logic data into the business layer interface respectively, and obtain first output data corresponding to the normal logic data and second output data corresponding to the special logic data; Judge whether the data presentation forms of the first output data and the second output data are the same. If so, determine that the business layer interface is normal; if not, determine that the business layer interface is abnormal.

2. The game interface testing method according to claim 1, wherein The extracting the normal data logic from the business layer interface and generating normal logic data according to the normal data logic includes: Split the business layer interface into a logic layer interface and a data layer interface; Extract first normal data logic from the logic layer interface and extract second normal data logic from the data layer interface; Generate first normal logic data according to the first normal data logic and generate second normal logic data according to the second normal data logic.

3. The game interface testing method according to claim 2, wherein The determining the special data logic according to the normal data logic and using a deep learning model to generate special logic data based on the special data logic includes: Determine first special data logic according to the first normal data logic, and use a deep learning model to generate first special logic data based on the first special data logic; Determine second special data logic according to the second normal data logic, and use a deep learning model to generate second special logic data based on the second special data logic.

4. A game interface testing method according to claim 3, characterized in that The inputting the normal logic data and the special logic data into the business layer interface respectively, and obtaining first output data corresponding to the normal logic data and second output data corresponding to the special logic data includes: Input the first normal logic data and the first special logic data into the logic layer interface respectively, and obtain first logic layer output data corresponding to the first normal logic data and second logic layer output data corresponding to the first special logic data; Input the second normal logic data and the second special logic data into the data layer interface respectively, and obtain first data layer output data corresponding to the second normal logic data and second data layer output data corresponding to the second special logic data.

5. A game interface testing method according to claim 4, characterized in that The judging whether the data presentation forms of the first output data and the second output data are the same. If so, determine that the business layer interface is normal; if not, determine that the business layer interface is abnormal includes: Judge whether the data presentation forms of the first logic layer output data and the second logic layer output data are the same. If so, determine that the logic layer interface is normal; if not, determine that the logic layer interface is abnormal; Determine whether the data presentation forms of the output data of the first data layer and the output data of the second data layer are the same. If so, determine that the data layer interface is normal; if not, determine that the data layer interface is abnormal.

6. A game interface testing system, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, the following steps are implemented: Extract the normal data logic from the service layer interface and generate normal logic data according to the normal data logic; Determine the special data logic according to the normal data logic and generate special logic data based on the special data logic using a deep learning model; Input the normal logic data and the special logic data into the service layer interface respectively to obtain the first output data corresponding to the normal logic data and the second output data corresponding to the special logic data; Determine whether the data presentation forms of the first output data and the second output data are the same. If so, determine that the service layer interface is normal; if not, determine that the service layer interface is abnormal.

7. A game interface testing system according to claim 6, characterized in that The step of extracting the normal data logic from the service layer interface and generating normal logic data according to the normal data logic includes: Split the service layer interface into a logic layer interface and a data layer interface; Extract the first normal data logic from the logic layer interface and extract the second normal data logic from the data layer interface; Generate the first normal logic data according to the first normal data logic and generate the second normal logic data according to the second normal data logic.

8. A game interface test system according to claim 7, wherein The step of determining the special data logic according to the normal data logic and generating special logic data based on the special data logic using a deep learning model includes: Determine the first special data logic according to the first normal data logic and generate the first special logic data based on the first special data logic using a deep learning model; Determine the second special data logic according to the second normal data logic and generate the second special logic data based on the second special data logic using a deep learning model.

9. The game interface test system according to claim 8, characterized in that The step of inputting the normal logic data and the special logic data into the service layer interface respectively to obtain the first output data corresponding to the normal logic data and the second output data corresponding to the special logic data includes: Input the first normal logic data and the first special logic data into the logic layer interface respectively to obtain the first logic layer output data corresponding to the first normal logic data and the second logic layer output data corresponding to the first special logic data; Input the second normal logic data and the second special logic data into the data layer interface respectively to obtain the first data layer output data corresponding to the second normal logic data and the second data layer output data corresponding to the second special logic data.

10. A game interface testing system according to claim 9, characterized in that The step of determining whether the data presentation forms of the first output data and the second output data are the same. If so, determine that the service layer interface is normal; if not, determine that the service layer interface is abnormal includes: Determine whether the data representation forms of the output data of the first logic layer and the output data of the second logic layer are consistent. If so, determine that the logic layer interface is normal; if not, determine that the logic layer interface is abnormal. Determine whether the data representation forms of the output data of the first data layer and the output data of the second data layer are consistent. If so, determine that the data layer interface is normal; if not, determine that the data layer interface is abnormal.

Images