Business processing method, device and equipment of game scene, and storage medium

Abstract

Embodiments of the application disclose a game scene service processing method and device, equipment and a storage medium, which are applied to the fields of artificial intelligence, block chain and traffic, and the method comprises the following steps: when a game scene meets a game service recommendation condition, triggering a scene recommendation request; based on the scene recommendation request, obtaining a set of scene service parameters matched with the game scene; based on the set of scene service parameters, screening a target service device matched with the scene recommendation request in a service device cluster, extracting scene feature data matched with the scene recommendation request through the target service device; and according to scene recommendation data generated by the target service device processing the scene feature data, performing game service recommendation in the game scene. Through the application, the development efficiency of game service can be improved.

Description

Technical Field

[0001] This application relates to fields such as artificial intelligence, blockchain, and cloud technology, and in particular to a business processing method, apparatus, device, and storage medium for a game scenario. Background Technology

[0002] With the continuous development of internet technology, the gaming industry has also gradually developed, and the types of game applications are increasing, such as single-player games and multiplayer games. At the same time, as gamers' demands for game applications increase, the development efficiency of game services has also improved. However, when developing new game services, it is necessary to completely copy the code corresponding to the original game service, modify the data in the copied code to generate the new game service, resulting in relatively low development efficiency. Summary of the Invention

[0003] This application provides a business processing method, apparatus, device, and storage medium for game scenarios, which can improve the development efficiency of game services.

[0004] This application provides a business processing method for a game scene, including:

[0005] When a game scene meets the game business recommendation criteria, a scene recommendation request is triggered.

[0006] Based on the scenario recommendation request, obtain the scenario business parameter set that matches the game scenario;

[0007] Based on the scenario service parameter set, target service devices that match the scenario recommendation request are selected from the service device cluster;

[0008] The scene feature data matching the scene recommendation request is extracted from the target service device;

[0009] Based on the scene recommendation data generated by the target service device processing the scene feature data, game service recommendations are made in the game scene.

[0010] One embodiment of this application provides a business processing device for a game scene, including:

[0011] The trigger module is used to trigger a scene recommendation request when the game scene meets the game business recommendation conditions;

[0012] The acquisition module is used to acquire a set of scene business parameters that match the game scene based on the scene recommendation request;

[0013] The filtering module is used to filter target business devices in the business device cluster that match the scenario recommendation request based on the scenario business parameter set.

[0014] The extraction module is used to extract scene feature data that matches the scene recommendation request through the target service device;

[0015] The generation module is used to recommend game services in the game scene based on the scene recommendation data generated by the target service device through processing the scene feature data.

[0016] One embodiment of this application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described.

[0017] One embodiment of this application provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the method described.

[0018] One aspect of this application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described.

[0019] In this application, when a game scene meets the game service recommendation conditions, the game service platform triggers a scene recommendation request. Based on this request, it obtains a set of scene service parameters matching the game scene. According to these parameters, it filters target service devices matching the request from the service device cluster, enabling the scene recommendation request to be routed to the target device. Furthermore, it allows for the routed distribution of different scene recommendation requests to different service devices, achieving distributed processing of scene recommendation requests. Further, it can extract scene feature data matching the request from the target service device. Based on the scene recommendation data generated by processing this feature data, game service recommendations are performed within the game scene, enabling online recommendation throughout the entire game process—that is, online processing of game services for multiple game scenes. Therefore, the game service platform implements the service routing process for scene recommendation requests. This routing process determines which service device will handle the request, and the target service device handles the request. The routing and processing processes are independent of each other, achieving decoupling and facilitating efficient processing of different game services. Meanwhile, since the business diversion process and the business processing process are independent of each other, when a user has new game business requirements, only the new game business processing algorithm needs to be added to the target business device, without needing to process the business diversion process, which helps to improve the development efficiency of game business. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the architecture of a business processing system for a game scene provided in this application;

[0022] Figure 2a This is a schematic diagram of the interaction scenario between various devices in a business processing system for a game scene provided in this application;

[0023] Figure 2b This is a schematic diagram of the interaction scenario between various devices in a business processing system for a game scene provided in this application;

[0024] Figure 3 This is a flowchart illustrating a business processing method for a game scenario provided in this application;

[0025] Figure 4 This is a schematic diagram illustrating an application scenario of a traffic diversion strategy provided in this application;

[0026] Figure 5 This is a schematic diagram illustrating the interaction scenario between the target resource management device and the target business device provided in this application;

[0027] Figure 6 This is a flowchart illustrating the business processing method for the game scenario provided in this application;

[0028] Figure 7 This is a flowchart illustrating a business processing method for a game scenario provided in this application;

[0029] Figure 8 This is a schematic diagram of the scenario for training the initial business processing model provided in this application;

[0030] Figure 9 This is a schematic diagram of the structure of a business processing device for a game scene provided in an embodiment of this application;

[0031] Figure 10 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0032] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0033] A game business processing solution comprises an engineering component and a business processing procedure. The engineering component refers to the process of providing basic services such as player referral mechanisms, while the business processing procedure refers to the process of providing game scene services such as recommending game items, matching players, and matching game scenes. The business processing procedure can also be referred to as the business processing algorithm. If the engineering component and the business processing procedure are coupled together, developing new game services requires completely copying the code corresponding to the original game service, modifying the data in the copied code, and generating the new game service, resulting in low development efficiency. Here, "new game service" can refer to deploying new game features within the original game application, or it can refer to deploying a completely new game application.

[0034] Addressing the issue of low development efficiency in game services, this application decouples the engineering portion from the business processing process (i.e., the business processing algorithm portion). Specifically, it implements the business routing process within the engineering portion of the game service through network management devices and target resource management devices, while the business processing process is handled by the target service device. This facilitates efficient handling of different game services. Furthermore, when developing new game services, only the new processing algorithm needs to be added to the target service device; the business routing process does not require rework, further improving development efficiency. During business processing, the target service device can employ machine learning techniques from artificial intelligence to process game services, improving accuracy and efficiency, and providing personalized gaming experiences for virtual game objects.

[0035] Understandably, this application separates the engineering portion into a game business platform, providing general basic services and traffic distribution mechanisms. This allows for efficient handling of numerous different game scenarios. Basic services include business characteristic data management, traffic distribution mechanisms, engineering indicator management, and more. In other words, the game business platform corresponding to the engineering portion of this application can provide general basic services and traffic distribution mechanisms for different game functions within the same game application; simultaneously, different game applications can share these general basic services and traffic distribution mechanisms. Furthermore, by deploying different game business-specific processing algorithms on the business devices, refined game operation can be achieved, providing personalized experiences for gamers and resulting in better performance. For example, matching suitable gamers to a game, recommending appropriate game items (such as in-game props) at the right time, and adjusting game difficulty at the right time. These processing algorithms are derived from game player data modeling experiments and deployed on an experimental platform separated from the engineering portion, significantly improving the development and operational efficiency of the experiments.

[0036] Understandably, 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, operating / interactive systems, and mechatronics. AI software technologies mainly include computer vision, speech processing, natural language processing, as well as machine learning / deep learning, autonomous driving, and intelligent transportation.

[0037] Machine learning (ML) in artificial intelligence is a multidisciplinary field involving probability theory, statistics, approximation theory, convex analysis, and algorithm complexity theory, among others. 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, inductive learning, and instructional learning.

[0038] To facilitate a clearer understanding of this application, we first introduce the business processing system for the game scenario that implements the business processing method of this application. This business processing system can also be referred to as a game business platform, such as... Figure 1 As shown, the business processing system of this game scene includes a gateway layer 10, a resource management layer 20, and a business layer.

[0039] It is understood that gateway layer 10 includes one or more network management devices, which may be gateway devices. This application does not limit the number of network management devices, such as... Figure 1 In this system, network management devices include network management device 101, network management device 102, ..., network management device 150, etc. These network management devices are interconnected via a network to facilitate data exchange. Each network management device can serve as an external service device in the business processing system, receiving scenario recommendation requests for game services from terminal devices, dispatching the processing requests to downstream service devices, and then sending the processing results back to the terminal devices after processing the game services.

[0040] Understandably, when there is only one network management device, this single device can handle scene recommendation requests from one or more game applications; when there are at least two network management devices, these at least two devices can handle scene recommendation requests from one or more game applications. Different network management devices can be used to handle scene recommendation requests from different game applications, or different network management devices can be used to handle scene recommendation requests sent by terminal devices belonging to different locations, or different network management devices can be used to handle scene recommendation requests from virtual game objects belonging to different skill levels, and so on. For example, such as... Figure 1 As shown, network management device 101 can be used to process scene recommendation requests corresponding to game application 1, network management device 102 can be used to process scene recommendation requests corresponding to game application 2, ..., network management device 150 can be used to process scene recommendation requests corresponding to game application 3.

[0041] The resource management layer 20 is composed of a cluster of resource management devices, which includes multiple resource management devices. This application does not limit the number of resource management devices. Figure 1 In this context, resource management devices include resource management devices 201, 202, ..., 280, etc., which are connected via a network to facilitate data exchange. Each resource manager is associated with a corresponding business layer cluster of business devices, and each business device cluster includes one or more business devices. For example... Figure 1As shown, resource management device 201 is associated with service device cluster 31, which includes service devices 301 and 302. Service devices 301 and 302 are connected via a network to facilitate data exchange between them. Resource management device 202 is associated with service device cluster 32, which includes service devices 303 and 304. Service devices 303 and 304 are connected via a network to facilitate data exchange between them. Resource management device 280 is associated with service device cluster 33, which includes service devices 305 and 306. Service devices 305 and 306 are connected via a network to facilitate data exchange between them.

[0042] Understandably, a resource management device and its associated cluster of business devices can be considered a set. Each set can be deployed independently in terms of hardware and software, and each set is logically independent, possessing its own independent functionality. For example, ... Figure 1 In this setup, resource management device 201 and business device cluster 31 constitute set1, resource management device 202 and business device cluster 32 constitute set2, and resource management device 203 and business device cluster 33 constitute set3. Set1 is used to provide business processing services for game scenes for virtual game objects whose game accounts belong to [0, 1000], set2 is used to provide business processing services for game scenes for virtual game objects whose game accounts belong to [1001, 2000], and so on.

[0043] Understandably, the different sets in the business processing system of this game scenario also have disaster recovery capabilities. For example, when all the business devices in set1 are unable to work properly, such as when the resource management device of set1 experiences a network failure, or when all the business devices in set1 malfunction, the game business that set1 needs to process can be migrated to set2. Set2 will then handle the game business that set1 needs to process, and at the same time, set2 will also handle the game business that set2 itself needs to process, thus improving device utilization. Similarly, the business devices within a set also have disaster recovery capabilities. For example, if business device 301 in set1 is unable to work properly, the game business that business device 301 needs to process can be migrated to business device 302. Business device 302 will then handle the game business that business device 301 needs to process, and at the same time, business device 302 will also handle the game business that business device 302 itself needs to process, ensuring that the game business is processed normally and improving device utilization.

[0044] Understandably, each set includes a resource management device and multiple service devices. The resource manager selects service devices from its associated service device cluster to handle game services. For example, the resource manager can select service devices from its associated service device cluster to handle game services based on at least one or a combination of traffic splitting and load balancing strategies. Traffic splitting strategy refers to the strategy of allocating recommendation requests for different scenarios to different service devices. For example, a traffic splitting strategy could be based on the object attribute information of a virtual game object in a game scenario, allocating the scenario recommendation request for that virtual game object to the corresponding service device. This object attribute information includes one or more of the following: the virtual game object's game account, the game scenario's scene identifier, the location information of the terminal device corresponding to the virtual game object, the service identifier corresponding to the virtual game object, and the game rank of the virtual game object. Load balancing strategy can be based on the current load of the service devices in the service device cluster, selecting service devices to handle scenario recommendation requests. For example, allocating scenario recommendation requests to service devices with relatively low current load. Load can refer to the amount of game services (scene recommendation requests to be processed) that a service device currently needs to handle.

[0045] Understandably, the business equipment is equipped with at least one business processing algorithm for handling scene recommendation requests. This algorithm can include one or more of the following: scene matching algorithm, item recommendation algorithm, and player matching algorithm. A scene matching algorithm can refer to an algorithm used to match game scenes, such as practice scenes and real-world battle scenes. An item recommendation algorithm can refer to an algorithm used to recommend game items to virtual game objects. A player matching algorithm refers to an algorithm used to match virtual game objects with opposing players (game teams). These business processing algorithms can be presented through a business processing model, which can be a machine learning model, including Naive Bayes models, decision tree models, nearest neighbor algorithms, support vector machines (SVM), logistic regression models, etc. Understandably, based on the number of players, game scenes include single-player game scenes, multiplayer game scenes, etc.; based on the game area, game scenes can include indoor scenes or outdoor scenes, etc.

[0046] It is understood that resource management devices and business devices can be independent physical servers, server clusters or distributed systems consisting of at least two physical servers, or cloud servers providing 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, Content Delivery Network (CDN), and big data and artificial intelligence platforms. Terminal devices can specifically refer to in-vehicle terminals, smartphones, tablets, laptops, desktop computers, smart speakers, speakers with screens, smartwatches, etc., but are not limited to these. Various terminal devices and servers can be directly or indirectly connected via wired or wireless communication. The number of terminal devices and servers can be one or at least two; this application does not impose any limitations. It should be noted that when the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use, and processing of user information such as object attribute information of virtual game objects, game behavior data, and scene feature data associated with scene recommendation requests must comply with the relevant laws, regulations, and standards of the relevant countries and regions. In other words, a computer device can only acquire user information such as object attribute information, game behavior data, and scene feature data associated with scene recommendation requests when it obtains authorization information from the user corresponding to the virtual game object regarding the aforementioned information. For example, the computer device can display a permission prompt interface in the game application's interface. This permission prompt interface is used to inform the user that the device is currently collecting the user's object attribute information. Only after the user confirms the permission prompt interface will the process of acquiring the user's object attribute information begin; otherwise, it will end.

[0047] based on Figure 1 The business processing system for the game scenario shown can be called a full-process online game recommendation system. This system can be used to implement scenario recommendation requests such as adjusting game difficulty levels, matching game scenarios, recommending game items, and matching players. When this full-process online game recommendation system executes these scenario recommendation requests... Figure 1 Interaction scenarios between various devices in the system, such as Figure 2a As shown, Figure 2a The network management device 10a in the text can refer to Figure 1 Any network management device in the gateway layer, Figure 2a Resource management equipment 20b in the text can refer to Figure 1 Resource management equipment 201, Figure 2a The resource management device 21b in the text can refer to Figure 1 Resource management equipment 202 in the middle, Figure 2aThe resource management device 22b in the text can refer to Figure 1 Resource management equipment 280. Figure 2a Business equipment 30c belongs to Figure 1 Central business equipment cluster 31, Figure 2a The business equipment 31c in the middle belongs to Figure 1 Central business equipment cluster 32, Figure 2a The business equipment 32c in the middle belongs to Figure 1 Central business equipment cluster 33. Figure 2a Terminal device 1, terminal device 2, and terminal device 3 in the text are the terminal devices corresponding to virtual game object 1, virtual game object 2, and virtual game object 3, respectively.

[0048] like Figure 2aAs shown, when virtual game object 1 needs to match a game scene, such as after the virtual game object logs into the game application and a trigger operation of the virtual game object on the "Start Game" option in the game application is detected, it is determined that the game scene meets the game service recommendation conditions. Terminal device 1 can generate a scene recommendation request 1 for matching the game scene and send the scene recommendation request 1 to network management device 10a. When network management device 10a receives the scene recommendation request 1 for matching the game scene from terminal device 1, it obtains the scene service parameter set for matching the game scene based on the scene recommendation request 1. The scene service parameter set includes one or more of the following: the traffic splitting keyword 1, the traffic splitting operation method 1, the traffic splitting parameters, and the scene identifier of the game scene. The traffic splitting keyword 1 can be generated based on the object attribute information of virtual game object 1 in the game scene. Here, the object attribute information of virtual game object 1 includes one or more of the following: the game account of virtual game object 1, the location information of terminal device 1 corresponding to virtual game object 1, the service identifier corresponding to virtual game object 1, and the game rank of virtual game object 1. The network management device 10a selects a resource management device from the resource management device cluster that matches the scene recommendation request, thus routing the scene recommendation request to the matching resource management device. For example, the object attribute information (i.e., the routing keyword) of the virtual game object 1 includes a service identifier 1, which reflects that the game service requested by the terminal device 1 is a matching game scene. Assuming that the service device cluster associated with resource management device 20b in the resource management device cluster is a service device cluster used to process matching game scenes, network management device 10a can determine resource management device 20b as the resource management device that matches the object attribute information and send the object attribute information of virtual game object 1 to resource management device 20b. Furthermore, resource management device 20b can select a service processing device that matches the scene recommendation request from its associated service device cluster and send the scene recommendation request to the matched service device, thus routing scene recommendation request 1 to the matching service device. For example, the object attribute information (i.e., the routing keyword) of the virtual game object 1 includes the game account of the virtual game object 1. Assuming that the set of game accounts corresponding to the game services processed by business device 30c in the business device cluster associated with resource management device 20b includes the game account of the virtual game object 1, resource management device 20b can identify business device 30c as the business device matching scene recommendation request 1 and send the object attribute information of the virtual game object 1 to business device 30c. Business device 30c includes scene matching algorithm 1, scene matching algorithm 2, ...; business device 30c can select one scene matching algorithm from these algorithms, match the game scene with scene recommendation request 1 according to the selected algorithm, execute scene recommendation request 1, and return the execution result to terminal device 1.Here, the virtual game object can refer to a game player who sends a scene recommendation request. It is evident that the business processing system for this game scene can provide online services to match virtual game objects with game scenes.

[0049] Similarly, such as Figure 2aAs shown, when virtual game object 2 needs to recommend game items, such as after the virtual game object logs into the game application, and when it is determined that the game scene of the virtual game object meets the game item recommendation conditions, terminal device 2 (or the application server corresponding to the game application) can generate a scene recommendation request 2 for recommending game items and send the scene recommendation request 2 to network management device 10a. Network management device 10a can obtain a scene service parameter set matching the game scene. The scene parameter service set includes one or more of the following: the traffic splitting keyword 2, traffic splitting operation method 2, traffic splitting parameters, and the scene identifier of the game scene. The traffic splitting keyword 2 can be generated based on the object attribute information of virtual game object 2 in the game scene. The object attribute information of virtual game object 2 includes one or more of the following: the game account of virtual game object 2, the location information of the terminal device 2 corresponding to virtual game object 2, the service identifier corresponding to virtual game object 2, and the game rank of virtual game object 2. When network management device 10a receives a scene recommendation request 2 from terminal device 2 regarding recommended game items, it selects a resource management device from the resource management device cluster that matches the scene recommendation request 2, thus routing the scene recommendation request 2 to the matching resource management device. For example, the object attribute information of the virtual game object 2 includes a service identifier 2, which reflects that the game service requested by terminal device 2 is a recommended game item. Assuming that the service device cluster associated with resource management device 21b in the resource management device cluster is a service device cluster used to process recommended game items, network management device 10a can determine resource management device 21b as the resource management device that matches the scene recommendation request 2 and send the scene recommendation request 2 to resource management device 21b. Furthermore, resource management device 21b can select a service processing device that matches the scene recommendation request 2 from its associated service device cluster and send the scene recommendation request 2 to the matched service device, thus routing the scene recommendation request 2 to the matching service device. For example, the object attribute information (i.e., the routing key 2) of the virtual game object 2 includes the game account of the virtual game object 2. Assuming that the set of game accounts corresponding to the game business processed by the business device 31c in the business device cluster associated with the resource management device 21b includes the game account of the virtual game object 2, the resource management device 21b can identify the business device 31c as the business device that matches the scene recommendation request 2 of the virtual game object 2, and send the scene recommendation request 2 of the virtual game object 2 to the business device 31c.The service device 31c includes item recommendation algorithm 1, item recommendation algorithm 2, ...; the service device 31c can select one of these item recommendation algorithms, recommend game items for the virtual game object 2 according to the selected item recommendation algorithm, execute scene recommendation request 2, and return the execution result to the terminal device 2. It can be seen that the service processing system of this game scene can provide online service for recommending game items for virtual game objects.

[0050] Similarly, such as Figure 2a and Figure 2bAs shown, when virtual game object 3 needs to match with players, if the terminal device 3 detects the virtual game object's triggered operation on the game application's battle option after logging into the game application, the terminal device 3 can generate a scene recommendation request 3 for matching with players and send the scene recommendation request 3 to the network management device 10a. The scene recommendation request 1 obtains the scene service parameter set for matching the game scene. This scene service parameter set includes one or more of the following: the game scene's traffic splitting keyword 3, traffic splitting operation method 3, traffic splitting parameters, and the game scene's scene identifier. The traffic splitting keyword 3 can be generated based on the object attribute information of the virtual game object 3 in the game scene. Here, the object attribute information of the virtual game object 3 includes one or more of the following: the virtual game object 3's game account, the location information of the terminal device 3 corresponding to the virtual game object 3, the service identifier corresponding to the virtual game object 3, and the game rank of the virtual game object 3. The object attribute information of virtual game object 3 includes one or more of the following: game account of virtual game object 3, location information of terminal device 3 corresponding to virtual game object 3, service identifier corresponding to virtual game object 3, and game rank of virtual game object 3. When network management device 10a receives a scene recommendation request 3 from terminal device 3 regarding matching players, it selects a resource management device from the resource management device cluster that matches the scene recommendation request 3, thereby routing the scene recommendation request 3 to the matching resource management device. For example, the object attribute information of virtual game object 3 (i.e., routing keyword 3) includes a service identifier 3, which reflects that the game service requested by terminal device 3 is matching players. Assuming that the service device cluster associated with resource management device 22b in the resource management device cluster is a service device cluster used to handle matching players, network management device 10a can determine resource management device 22b as the resource management device that matches the scene recommendation request 3 and send the scene recommendation request 3 to resource management device 22b. Furthermore, resource management device 22b can select a service processing device matching the scenario recommendation request 3 from its associated service device cluster, and send the scenario recommendation request 3 to the matched service device, thus achieving the diversion of scenario recommendation request 3 to the matched service device. For example, the object attribute information of the virtual game object 3 includes the game account of the virtual game object 3. Assuming that the set of game accounts corresponding to the game service processed by service device 32c in the service device cluster associated with resource management device 22b includes the game account of the virtual game object 3, resource management device 22b can determine service device 32c as the service device matching the scenario recommendation request 3, and send the scenario recommendation request 3 to service device 32c.The service device 32c includes player matching algorithm 1, player matching algorithm 2, ...; these player matching algorithms have corresponding relationships with different game applications, such as player matching algorithm 1 corresponding to game application 1, player matching algorithm 2 corresponding to game application 2, ... In this case, the service device 32c can use the player matching algorithm that has a corresponding relationship with the game application that sent the scenario recommendation request 3 as the player matching algorithm for matching scenario recommendation request 3. Alternatively, these player matching algorithms have corresponding relationships with different matchmaking player services, such as player matching algorithm 1 corresponding to 1V(pair) 1 matchmaking player service, player matching algorithm 2 corresponding to 3V3 matchmaking player service, ... In this case, the service device 32c can use the player matching algorithm that has a corresponding relationship with the service identifier of the matchmaking player service in the scenario recommendation request 3 as the player matching algorithm for matching scenario recommendation request 3. Furthermore, the service device 32c can acquire a candidate player set, which includes at least two candidate players, and acquire the service feature data of each candidate player and the service feature data of the virtual game object. It can also acquire the similarity between the service feature data of the virtual game object and the service feature data of each candidate player, and determine the matching degree between the service feature data of the virtual game object and each candidate player based on this similarity. For example... Figure 2bAs shown, the candidate player set includes candidate player 1, candidate player 2, ... The similarity score 1 between the business characteristic data of candidate player 1 and the business characteristic data of the virtual game object is obtained. Based on this similarity score 1, the matching score 1 between candidate player 1 and the virtual game object is determined. Similarly, the similarity score 2 between the business characteristic data of candidate player 2 and the business characteristic data of the virtual game object is obtained. Based on this similarity score 2, the matching score 2 between candidate player 2 and the virtual game object is determined, and so on. Then, candidate players with a matching score greater than a matching score threshold are selected from the candidate player set as the players to compete against in the virtual game object. Assuming this matchmaking service is a 3V3 matchmaking service, service device 32c can select 5 candidate players with a matching score greater than the matching score threshold from the candidate player set as the players to compete against in the virtual game object. Two of these 5 players form one team with the virtual game object player, and the other 3 players form another team. The matching players that business device 32c can match are returned to terminal device 3. Terminal device 3 can then display the matching results in the game interface 33d. These results include players 1 and 2 who belong to the same team as the virtual game object, and players 3, 4, and 5 who belong to the same team as the virtual game object. Therefore, the business processing system for this game scenario can provide online matching services for virtual game objects. It is understandable that the traffic splitting operation method includes at least one of modulo operation and hash operation followed by modulo operation. The traffic splitting parameters are fixed values. Specifically, different game scenarios can have the same traffic splitting operation method and parameters. In this case, different game scenarios share a single traffic splitting mechanism (i.e., traffic splitting strategy), which can improve traffic splitting efficiency. Alternatively, different game scenarios can have different traffic splitting operation methods and parameters, thus providing a personalized traffic splitting mechanism for different game scenarios.

[0051] In summary, by using network management and resource management devices to implement the service routing process for game services, and by using service devices to implement the service processing process for game services, the decoupling between the engineering part (service routing process) and the service processing process is achieved. When developing new game services, only the new game service processing algorithm needs to be added to the service devices, without needing to process the service routing process, which helps to improve the development efficiency of game services.

[0052] Further, please see Figure 3 This is a flowchart illustrating a business processing method for a game scene provided in an embodiment of this application. Figure 3 As shown, this method can be derived from... Figure 1 The game business platform in the context of the game scenario. The business processing method for this game scenario may include the following steps S101–S105:

[0053] S101. When a game scene meets the game business recommendation conditions, the game business platform triggers a scene recommendation request.

[0054] In this embodiment of the application, when a virtual game object logs into a game application, the terminal device corresponding to the virtual game object or the application service device of the game application can detect whether the game scene corresponding to the virtual game object meets the game service recommendation conditions. When the game scene meets the game service recommendation conditions, a scene recommendation request is triggered and the scene recommendation request is sent to the network management device in the game service platform.

[0055] Understandably, if the game difficulty of the game scene meets the adjustment conditions, then the game scene is determined to meet the game service recommendation conditions, and a scene recommendation request is generated to instruct the adjustment of the game difficulty level of the game scene; if the game scene meets the game item recommendation conditions, then the game scene is determined to meet the game service recommendation conditions, and a scene recommendation request is generated to instruct the recommendation of game items to the game scene; if an instruction to match game players for the game scene is received, then the game scene is determined to meet the game service recommendation conditions, and a scene recommendation request is generated to instruct the matching of game players to virtual game objects in the game scene.

[0056] For example, after a virtual game object logs into the game application, the application server can obtain the first game behavior data of the virtual game. This first game behavior data includes the current game difficulty level of the virtual game object and the time required to complete a game at the current difficulty level. If the time required to complete a game at the current difficulty level is less than a time threshold, it is determined that the game difficulty level of the virtual game object's game scene meets the adjustment conditions. Based on the current game difficulty level of the virtual game object's game scene in the first game behavior data, a scene recommendation request is generated. This scene recommendation request is used to instruct the adjustment of the virtual game object's game difficulty level in the game scene, such as instructing the increase of the virtual game object's game difficulty level in the game scene. This facilitates the game service platform in this application to provide online adjustment services for the game difficulty of virtual game objects.

[0057] For example, after the virtual game object logs into the game application, the application server can obtain second game behavior data of the virtual game. This second game behavior data includes the virtual game object's winning probability in the game scene, the virtual game object's remaining game assets in the game application, and the game items required for the game scene, etc. Game assets can refer to resources used to purchase game items, such as game currency. If the winning probability is less than a probability threshold and the amount of remaining game assets is greater than an asset amount threshold, the application server can determine that the virtual game object's game scene meets the game item recommendation conditions. Based on the game items required for the game scene in the second game behavior data, the application server generates a scene recommendation request, which is used to instruct the virtual game object to be recommended game items. This is beneficial for the game service platform in this application to provide online game item recommendation services for virtual game objects. For another example, after the virtual game object logs into the game application, the application server can obtain third game behavior data of the virtual game. This third game behavior data is used to reflect the detected triggering operation of the virtual game object on the game application's battle options, and the type of battle options. For example, the battle options may include 1v1 or 3v3 battle options. After the application server obtains the third-party game behavior data, it can generate a scene recommendation request. The scene recommendation is used to instruct the virtual game object to be assigned a player to battle. This is beneficial for the game service platform in this application to provide online matchmaking services for virtual game objects.

[0058] S102. Based on the scenario recommendation request, the game business platform obtains the scenario business parameter set that matches the game scenario.

[0059] In this application, the network management device in the game service platform obtains a set of scene service parameters matching the game scene based on the scene recommendation request. This set of scene service parameters includes one or more of the following: the game scene's traffic splitting keyword, traffic splitting calculation method, traffic splitting parameters, and the game scene's scene identifier. The traffic splitting keyword can be generated based on the object attribute information of the game player in the game scene, or the traffic splitting keyword refers to the object attribute information of the virtual game object. The object attribute information of the virtual game object includes one or more of the following: the virtual game object's game account, the location information of the terminal device corresponding to the virtual game object, the service identifier corresponding to the virtual game object, and the game rank of the virtual game object 2.

[0060] S103. Based on the scenario service parameter set, the game service platform filters target service devices in the service device cluster that match the scenario recommendation request.

[0061] In this application, the network management device of the game service platform filters the target service device that matches the scenario recommendation request in the service device cluster based on the scenario service parameter set, so as to realize the diversion of the scenario recommendation request to the target service device.

[0062] Understandably, when the game service platform does not include a resource manager, the network management device of the game service platform can directly filter target service devices matching the scenario recommendation request from the service device cluster based on the scenario service parameter set. When the game service platform does not include a resource manager, the network management device of the game service platform can filter target resource management devices matching the scenario recommendation request from the resource management device cluster based on the scenario service parameter set, and send the scenario recommendation request to the target resource management device, thereby diverting the scenario recommendation request to the target resource management device. Further, the target resource management device obtains the service device cluster associated with it, and based on the scenario service parameter set, filters target service devices matching the scenario recommendation request from the service device cluster associated with it.

[0063] Optionally, the process by which the target resource management device selects target business devices matching the scenario recommendation request from the business device cluster associated with the target resource management device based on the scenario business parameter set includes: obtaining the business device relationship tree corresponding to the target resource management device; the parent node of the business device relationship tree is the target resource management device, and the child nodes of the business device relationship tree are the business devices in the business device cluster associated with the target resource management device; traversing the business devices in the business device relationship tree according to the node path of the business node relationship tree; and selecting target business devices matching the scenario recommendation request from the business device relationship tree based on the scenario business parameter set.

[0064] The target resource management device can obtain its own business device relationship tree. The parent node of this relationship tree is the target resource management device, and its child nodes are the business devices in the business device cluster associated with it. The position of a business device within a child node (i.e., the traversal order) can be determined based on the frequency of its game processing or its performance parameters during a historical time period. The target resource management device can traverse the business devices in the relationship tree from top to bottom or from bottom to top along the node paths. Based on the scenario business parameter set, it can filter target business devices from the relationship tree that match the scenario recommendation request. By using the business device relationship tree, target business devices matching the scenario recommendation request can be quickly found, improving the efficiency of business device matching.

[0065] Optionally, the target resource management device may use any one or a combination of the following four methods to determine the target service device for processing the target game service:

[0066] Method 1: The scenario's business parameter set includes the traffic splitting keyword, traffic splitting calculation method, and traffic splitting parameters corresponding to the game scenario. The traffic splitting calculation method includes modulo, whitelist calculation, hash-based modulo, etc. The key to traffic splitting is located in the configuration parameters, which have two parts: one part is the traffic splitting keyword generated based on the object attribute information of virtual game objects in the game scenario, and the other part is the return result of the traffic splitting strategy, which is a numerical value (called the traffic splitting value). The configuration parameters improve the flexibility of traffic splitting. For example, the hash-based modulo method can achieve orthogonal experimental traffic splitting by configuring different hash parameters; the numerical range is used to implement mutually exclusive traffic splitting, determining the numerical range where the returned value falls to obtain the hit partition. Specifically, the target resource management device in the game business platform performs calculations on the traffic splitting keyword and the traffic splitting parameters according to the traffic splitting calculation method to obtain the traffic splitting value for the game scenario; it then obtains the numerical range corresponding to the business device in the business device relationship tree; this numerical range is generated based on the traffic splitting keyword corresponding to the game scenario for which the business device has management authority and the traffic splitting parameters, i.e., this numerical range reflects the scenario recommendation requests managed by the business device. Furthermore, the relationship between the numerical ranges corresponding to the business devices in the business device relationship tree and the routing values ​​of the game scene object is determined. Business devices with an inclusion relationship between the numerical ranges in the business device relationship tree and the routing values ​​of the game scene are identified as target business devices matching the scene recommendation request. By using this scene business parameter set, the scene recommendation request for the virtual game object is routed to the target business device. This allows for the routing of different scene recommendation requests to different business devices, improving the processing efficiency of game services and enabling distributed processing of game services.

[0067] For example, in a game scenario's business processing system, network management devices, resource management devices, and business devices all store traffic splitting policies (i.e., traffic splitting mechanisms). These traffic splitting policies may be the same or different. Figure 4 As shown, the explanation assumes that the traffic splitting strategy is the same in all devices. Figure 4 The network management device 50a in the text can refer to Figure 1 Any network management device in the network, Figure 4 The resource management device 51a in the text can refer to Figure 1 Resource management devices within any set, Figure 4In this context, service device 52a can refer to any service device within the set corresponding to resource management device 51a. Network management device 50a, resource management device 51a, and service device 52a all store the same traffic splitting strategy. This strategy comprises a shuffle method, configuration parameters, and numerical ranges. It defines the partitions corresponding to network management device 50a, resource management device 51a, and service device 52a, respectively. Each partition has one or more numerical ranges, and the numerical ranges for different partitions are different. The numerical range of this device can be generated based on the game account, game rank, and location information of the corresponding terminal device of the virtual game object with management authority. For example, if the game account of the virtual game object with management authority in resource management device 51a is [0, 1000], and the configuration parameter is 100, then the numerical range [0, 99] is obtained by modulo operation on the game account of the virtual game object with management authority in resource management device 51a according to this configuration parameter. Therefore, the numerical range corresponding to resource management device 51a is [0, 99]. Specifically, the partition corresponding to network management device 50a is associated with a set, the partition corresponding to resource management device 51a is associated with a service device, and the partition corresponding to service device 52a is associated with a service processing algorithm.

[0068] From request processing to the underlying business processing algorithm, the request passes through different devices, requiring a stable traffic distribution strategy. For network management device 50a, traffic distribution involves querying the set (i.e., the matching resource management device) that matches the object attribute information of the virtual game object. For example, if the game service of a certain virtual game object needs to be processed by the business processing algorithm deployed in set1, then the processing request for that virtual game object can be routed to set1, meaning the scene recommendation request matches set1. For resource management device 51a, traffic distribution involves finding the business device that matches the scene recommendation request from the business device cluster associated with resource management device 51a. For instance, one (or a group of) business devices in the business device cluster associated with resource management device 51a might have a stable business processing algorithm deployed, while another (or a group of) business devices might have an unstable one. The resource manager can distribute the scene recommendation request to different business devices (or groups of business devices) based on the scene parameter business set. When a processing request is distributed to a business device group, a business device with a lower load can be selected from the group as the target business device for processing the game service. For service device 52a, traffic splitting involves finding a specific business processing algorithm within that device. For example, it might use configuration parameters to perform modulo operations on the game account of a virtual game object to obtain a traffic splitting value. Based on the value range matched by this splitting value, it might derive the business processing algorithm associated with each partition. All three traffic splitting mechanisms use the same configuration and algorithm structure, and the configuration can be refreshed at runtime to adjust the traffic splitting strategy.

[0069] For example, Figure 4 As shown, assuming the traffic splitting keyword is the game account of a virtual game object, the value range corresponding to resource management device 51a is [0, 99]. The business device cluster associated with resource management device 51a includes business device 52a and business device 53a. The partitions corresponding to business device 52a and business device 53a are partition 1 and partition 2, respectively. The value range corresponding to partition 1 is [0, 49], and the value range corresponding to partition 2 is [50, 99]. If the traffic splitting parameter is 100, resource management device 51a can use this parameter to perform modulo processing on the game account of the virtual game object to obtain the traffic splitting value, that is, the game account of the virtual game object mod 100 = the traffic splitting value, assuming the value is 13. The value range hit by this traffic splitting value is [0, 49], and the partition hit by the traffic splitting value is partition 1. Therefore, business device 52a can be identified as the target business device matching the scene recommendation request. Similarly, the target business device can use the same method to select the business processing algorithm for processing game business.

[0070] Understandably, when a network management device receives a scene recommendation request, it can obtain the scene service parameter set for that game scene based on the request. This parameter set includes the service identifier of the target game service corresponding to the scene recommendation request. The network management device can then determine the target service algorithm for processing the target game service (i.e., the scene recommendation request) based on this identifier, identify the service device with the target service algorithm as the target service device for processing the target game service, and identify the resource management device associated with that target service device as the target resource management device. Alternatively, the scene service parameter set may include the game rank of virtual game objects in the game scene. The device performs a calculation on the game rank according to the traffic distribution parameters in the traffic distribution strategy (e.g., modulo operation, or hash operation followed by modulo operation) to obtain a traffic distribution value. It then determines the value range of the resource management devices in the resource management device cluster that this value matches, and identifies the resource management devices corresponding to this range as the target resource management devices matching the scene recommendation request. Furthermore, it determines the value range of the service devices associated with the target resource management device that the traffic distribution value matches, and identifies the service devices corresponding to this range as the target service devices for processing the target game service.

[0071] Method Two: The scenario service parameter set includes the location information of the terminal devices corresponding to the virtual game objects in the game scene. The target resource management device can obtain the location set corresponding to the business devices in the service device relationship tree, which includes the location information of the terminal devices that the business device has management authority over in the service device relationship tree. Further, the relationship between the location information of the terminal devices corresponding to the virtual game objects and the location set corresponding to the business devices in the service device relationship tree is determined. If the relationship between the location set corresponding to the business devices and the location information of the terminal devices corresponding to the virtual game objects is a non-inclusion relationship, it indicates that the business device does not have management authority over the terminal devices corresponding to the virtual game objects; if the relationship between the location set corresponding to the business devices and the location information of the terminal devices corresponding to the virtual game objects is an inclusion relationship, it indicates that the business device has management authority over the terminal devices corresponding to the virtual game objects. Therefore, the target resource management device can determine the business devices whose relationship with the location set corresponding to the business device relationship tree and the location information of the terminal devices corresponding to the virtual game objects is an inclusion relationship as the target business devices matching the scenario recommendation request. By using the location information of the terminal device corresponding to the virtual game object, the scene recommendation request can be diverted to the target business device. This can improve the efficiency of game business processing and realize distributed processing of game business by diverting scene recommendation requests with corresponding terminal devices in different locations to different business devices.

[0072] For example, suppose the business device relationship tree includes business device 1 and business device 2. The location set corresponding to business device 1 includes cities A1 and A2, and the location set corresponding to business device 2 includes cities B1 and B2. That is, business device 1 is used to manage terminal devices of virtual game objects located in cities A1 and A2, and business device 2 is used to manage terminal devices of virtual game objects located in cities B1 and B2. If the target resource management device receives a scene recommendation request for virtual game object 1, and the location of the terminal device corresponding to virtual game object 1 belongs to city A1, then the target resource management device can determine that there is an inclusion relationship between the location information of the terminal device corresponding to virtual game object 1 and the location set of business device 1, and can identify business device 1 as the target business device that matches the scene recommendation request for virtual game object 1. Similarly, if the target resource management device receives a scene recommendation request for virtual game object 2, and the location of the terminal device corresponding to virtual game object 2 belongs to city B2, then the target resource management device can determine that there is an inclusion relationship between the location information of the terminal device corresponding to virtual game object 2 and the location set of business device 2, and can determine business device 2 as the target business device that matches the scene recommendation request of virtual game object 1.

[0073] Method 3: The target resource management device can filter business devices matching the scenario recommendation request from the business device relationship tree based on the scenario business parameter set. These can be candidate business devices, with one or more. When there is only one candidate business device, the target resource management device can identify it as the target business device matching the scenario recommendation request. When there are multiple candidate business devices, the target resource management device can determine the target business device matching the scenario recommendation request from among these candidate business devices according to a load balancing strategy. Specifically, when there are multiple candidate business devices, the target resource management device can obtain the workload corresponding to the pending game business of each candidate business device. If the workload of a candidate business device is large, allocating the target game business to that candidate business device would result in a long waiting time before execution. Conversely, if the workload of a candidate business device is small, allocating the target game business to that candidate business device would allow for execution without a long waiting time. Therefore, the target resource management device can identify candidate business devices with workloads less than a workload threshold as the target business devices matching the scenario recommendation request. The target resource management device can use a combination of traffic splitting and load balancing strategies to determine the target business device that matches the scenario recommendation request, which can improve the efficiency of business processing in game scenarios and realize distributed processing of game business.

[0074] Method 4: The scenario service parameter set includes the scenario identifier of the game scenario. The target resource management device can determine the target service processing model for handling the scenario recommendation request based on the scenario identifier. The service devices in the service device relationship tree that possess the aforementioned service processing model are selected as the target service devices matching the scenario recommendation request. By allowing the target resource management device to determine the target service device matching the scenario recommendation request based on the scenario identifier, the efficiency of game scenario service processing can be improved, enabling distributed processing of game services.

[0075] Understandably, the implementation method of the network management device in the game business platform to select the target resource management device that matches the scene recommendation request from the resource management device cluster based on the scene business parameter set can refer to the above implementation method of the target resource manager to obtain the target business device that matches the scene recommendation request from the business device cluster based on the scene business parameter set. The repeated parts will not be repeated.

[0076] Understandably, the target service equipment drives the operation of multiple services in a consistent manner, and these services can reuse the common capabilities provided by the target service equipment, improving service implementation efficiency and engineering quality. For example... Figure 5 As shown, the target business device 60b may include an architecture layer, a business layer, and an infrastructure layer. The framework layer includes a configurator and a master controller. The configurator is used to create business interaction units according to configuration files and register them with the business factory. The master controller is used for system initialization and main logic loop of the target business device. The infrastructure layer provides basic functions such as writing logs, reading configurations, recording statistics, message queues, timer queues, business management, and message dispatching. The business layer includes business interaction units, business model units, business interaction unit interfaces, and business model unit interfaces. The business interaction unit is responsible for external business interactions. The special logic of the business is isolated in the business interaction unit. It calls the infrastructure layer to prepare feature data and complete engineering requirements, and calls the business model unit to complete the core business requirements. The business model unit realizes the core business value. Specifically, it compiles the business processing model into a dynamic link library and inserts it into the business interaction unit at runtime. The dynamic link library implements two agreed functions, including a dynamic loading function (dlsym) and a model instance creation function. The target business device obtains the storage address of the target business processing model through the dynamic loading function and executes it. Create a model instance named bu_$(businessID)_create_model to obtain descriptive information about the business processing model, such as the model version, named bu_$(businessID)_get_version.

[0077] Optionally, determining the target business processing model for handling the scene recommendation request based on the scene identifier of the game scene may include: the target business device can obtain a business processing model library for handling the target game business, and business processing logs of the business processing models in the business processing model library within a historical time period; determine the business processing performance parameters of the business processing models in the business processing model library based on the business processing logs; and determine the business processing model for handling the scene recommendation request from the business processing model library based on the business processing performance parameters, as the target business processing model.

[0078] like Figure 5 As shown, the target business device 60b obtains a processing request from the target resource management device through the business interaction unit interface and retrieves the business processing model library associated with the target game business. This business processing model library can refer to a dynamic link library within the business interaction unit. In other words, the business processing model library is inserted into the target business device as a dynamic link library, enabling plug-and-play functionality and improving the ease of application of the business processing model. This business processing model library includes at least two business processing models, each with different business processing performance parameters. These performance parameters reflect the business processing capabilities of the business processing model and include one or more of the following: business processing stability, business processing accuracy, and business processing efficiency. Therefore, when the target service device 60b receives a processing request, it can obtain the service processing logs of the service processing models in the service processing model library within a historical time period. This historical time period can refer to the past week, the past month, etc. Based on these logs, the service processing performance parameters of the service processing models in the library are determined. Then, based on these performance parameters, the service processing model used to process the scenario recommendation request is selected from the library as the target service processing model. For example, if the scenario recommendation request is a service with high real-time requirements, such as a matching scenario service, a service processing model with high efficiency in the library can be selected as the target service processing model. Similarly, if the target game service requires high accuracy, such as a player matching service, a service processing model with high accuracy in the library can be selected as the target service processing model. By selecting the service processing model for the scenario recommendation request based on its performance parameters, the processing accuracy and efficiency of the target game service can be improved.

[0079] For example, such as Figure 6As shown, when the network management device receives a scene recommendation request, it can obtain a set of scene parameter services matching the game scene based on the request. Based on this set, it determines the target resource management device matching the request from the resource management device cluster and generates a routing request. This routing request instructs the target resource management device to query and obtain the target service device matching the scene recommendation request. When the target resource management device receives the scene recommendation request, it generates a routing response based on the target service device matching the scene service parameter set and sends this response to the network management device. The network management device then forwards the scene recommendation request to the target service device based on this response. When the target service device receives the scene recommendation request, it can extract scene feature data matching the request and the target service processing model for processing the request. It then calls this model to generate scene recommendation data based on the scene feature data and performs game service recommendations within the game scene based on this data.

[0080] Optionally, the scenario recommendation request in this application can be used to instruct on matching players, matching scenarios, and recommending game items, etc. The following explanation uses the scenario recommendation request as an example of instructing on matching players. In this case, the scenario recommendation data generated by the target service device based on the storage address and calling the target service processing model, according to the scenario feature data, includes: obtaining, through the target service device, the business feature data of candidate players regarding the game scenario from the candidate player set, and the business feature data of the virtual game object; the business feature data of the candidate players here includes the candidate players' historical behavior data regarding the game scenario and the candidate players' real-time business parameters regarding the game scenario, etc. For example, the historical behavior data includes the number of abnormal events that occurred during the player's battle, the number of wins and losses in the player's battle, the number of times the player participated in the battle, etc. Abnormal events include one or more of the following: the player or candidate player leaving the battle during the battle, and the report submitted by the player or candidate player during the battle. Real-time business parameters include one or more of the following: game difficulty, game duration, number of players, etc. Similarly, the business characteristic data of virtual game objects here includes historical behavior data of virtual game objects regarding game scenarios and real-time business parameters of virtual game objects regarding game scenarios, etc. For example, the historical behavior data includes the number of abnormal events that occurred during the virtual game object's battle, the number of wins and losses of the virtual game object during the battle, the number of times the virtual game object participated in the battle, etc. Abnormal events include one or more of the following: events such as players or virtual game objects leaving the battle during the battle, and reports submitted by players or virtual game objects during the battle. Real-time business parameters include one or more of the following: game difficulty, game duration, number of players, etc.

[0081] Furthermore, the target business processing model can be invoked to match and identify the business characteristic data of the virtual game object with the business characteristic data of the candidate players in the candidate player set, obtaining the matching degree between the virtual game object and the candidate players in the candidate player set. This matching degree reflects one or more of the following: the probability of abnormal events occurring during the battle between the virtual game object and the candidate players, the difference in combat ability between the virtual game object and the candidate players, etc. Abnormal events here can refer to events such as quitting the battle game during the battle game or reporting events occurring during the battle game. Therefore, the target business device can select candidate players with a matching degree greater than the matching degree threshold from the candidate player set as the battle players for the virtual game object, and generate scene recommendation data based on the battle players of the virtual game object. For example, the object attribute information of the battle players of the virtual game object can be determined as scene recommendation data. This enables refined game operation, matching suitable battle players for virtual game objects, and providing personalized game experiences for virtual game objects.

[0082] Optionally, the above-mentioned method of calling the target business processing model through the target business device to match and identify the business feature data of the virtual game object and the business feature data of the candidate players in the candidate player set to obtain the matching degree between the virtual game object and the candidate players in the candidate player set includes: the target business device can call the abnormal event identification layer of the target business processing model to identify abnormal events between the business feature data of the virtual game object and the business feature data of the candidate players in the candidate player set, and obtain the abnormal event probability. The abnormal event probability is used to reflect the probability of an abnormal event occurring when the virtual game object plays against the candidate players in the candidate player set.

[0083] Furthermore, the capability recognition layer of the target business processing model can be invoked to identify the virtual game object's combat ability based on its business characteristic data, and to identify the combat ability of the candidate players in the candidate player set based on their business characteristic data. Combat ability reflects the probability of the virtual game object (or candidate player) winning during a combat game. Then, the matching degree between the virtual game object and the candidate players in the candidate player set can be determined based on the probability of abnormal events, the virtual game object's combat ability, and the combat ability of the candidate players in the candidate player set. By determining the matching degree between the virtual game object and the candidate players based on their combat abilities, the candidate players' combat abilities, and the probability of abnormal events, the probability of abnormal events occurring during combat games can be reduced, as can the problem of excessive differences in combat ability between the player and the virtual game object be addressed, providing the virtual game object with a better combat game experience and a personalized gaming experience.

[0084] Optionally, determining the matching degree between the virtual game object and the candidate players in the candidate player set based on the abnormal event probability, the combat ability of the virtual game object, and the combat ability of the candidate players in the candidate player set includes: the target service device can determine the difference in combat ability between the virtual game object and the candidate players in the candidate player set, and perform a weighted summation of the difference in combat ability and the abnormal event probability to obtain the combat difference value between the virtual game object and the candidate players in the candidate player set. Here, the combat difference value is correlated with both the combat difference and the abnormal event probability; that is, the larger the combat difference and the abnormal event probability, the larger the combat difference value, and vice versa. Furthermore, the matching degree between the virtual game object and the candidate players in the candidate player set can be determined based on the combat difference value. There is a negative correlation between the combat difference value and the matching degree between the virtual game object and the candidate players in the candidate player set; that is, the larger the combat difference value between the virtual game object and the candidate player, the lower the matching degree; conversely, the smaller the combat difference value, the higher the matching degree. By determining the matching degree between the virtual game object and the candidate player based on the combat difference value, it is beneficial to reduce the probability of abnormal events during the game and to prevent one side's combat ability from completely overwhelming the other, thus matching the virtual game object with the optimal player.

[0085] S104. Extract scene feature data that matches the scene recommendation request through the target business device.

[0086] S105. Based on the scene recommendation data generated by the target service device through processing the scene feature data, recommend game services in the game scene.

[0087] In steps S104 and S105, the target resource management device can send a scene recommendation request to the target service device. The target service device can then query and extract scene feature data matching the scene recommendation request, generate scene recommendation data based on this data, and perform game service recommendations within the game scene based on this recommendation data. The scene feature data includes historical behavior data of objects within the game scene and real-time service parameters. For example, if the scene recommendation request is for matching players, the historical behavior data includes the number of abnormal events that occur during the game between the virtual game object and the candidate player, the number of wins and losses between the virtual game object and the candidate player, the number of times the virtual game object and the candidate player participate in the game, etc. Abnormal events include one or more of the following: the player or virtual game object leaving the game during the game, and the report submitted by the player or virtual game object during the game. Real-time service parameters include one or more of the following: the game difficulty, game duration, number of players, etc., of the virtual game object in the current game scene. Furthermore, by using the scene recommendation data generated by the target business device based on the scene feature data, game business recommendations can be made in the game scene. This can decouple the scene recommendation request diversion process from the business processing process and improve the processing efficiency of game business.

[0088] In this application, when a game scene meets the game service recommendation conditions, the game service platform triggers a scene recommendation request. Based on this request, it obtains a set of scene service parameters matching the game scene. According to these parameters, it filters target service devices matching the request from the service device cluster, enabling the scene recommendation request to be routed to the target device. Furthermore, it allows for the routed distribution of different scene recommendation requests to different service devices, achieving distributed processing of scene recommendation requests. Further, it can extract scene feature data matching the request from the target service device. Based on the scene recommendation data generated by processing this feature data, game service recommendations are performed within the game scene, enabling online recommendation throughout the entire game process—that is, online processing of game services for multiple game scenes. Therefore, the game service platform implements the service routing process for scene recommendation requests. This routing process determines which service device will handle the request, and the target service device handles the request. The routing and processing processes are independent of each other, achieving decoupling and facilitating efficient processing of different game services. Meanwhile, since the business diversion process and the business processing process are independent of each other, when a user has new game business requirements, only the new game business processing algorithm needs to be added to the target business device, without needing to process the business diversion process, which helps to improve the development efficiency of game business.

[0089] Further, please see Figure 7 This is a flowchart illustrating a business processing method for a game scene provided in an embodiment of this application. Figure 7 As shown, this method can be derived from... Figure 1 The game business platform in the context of the game scenario. The business processing method for this game scenario may include the following steps S201–S207:

[0090] S201. When a game scene meets the game business recommendation conditions, the game business platform triggers a scene recommendation request.

[0091] S202. Based on the scenario recommendation request, the game business platform obtains the scenario business parameter set that matches the game scenario.

[0092] S203. Based on the scenario service parameter set, the game service platform filters target service devices in the service device cluster that match the scenario recommendation request.

[0093] S204. Obtain the dynamic loading function corresponding to the business processing model library through the target business device.

[0094] S205. By calling the dynamic loading function through the target service device, the storage address of the target service processing model in the service processing model library is obtained.

[0095] Optionally, the target service device can train the initial service processing model to obtain the service processing model in the service processing model library in the following ways: (e.g.) Figure 8 As shown, the target service device can obtain the service feature data of the sample game object regarding the scene recommendation request, the labeled players of the sample game object, and the service feature data of the sample players in the sample player set regarding the scene recommendation request. The scene recommendation request is used to instruct the allocation of players to the sample game object, and the labeled players of the sample game object belong to the sample players in the sample player set. Further, the target service device can invoke the initial service processing model to perform matching and prediction based on the service feature data of the sample game object and the service feature data of the sample players in the sample player set, thereby obtaining the predicted players for the sample game object; these predicted players belong to the sample player set. If the predicted players and the labeled players are the same or similar, it indicates that the accuracy of the initial service processing model in matching players is relatively high; if the difference between the predicted players and the labeled players is large, it indicates that the accuracy of the initial service processing model in matching players is relatively low. Therefore, the target service device can adjust the initial service processing model based on the predicted and labeled players to obtain a service processing model for handling the target game's business. This model is then added to the service processing model library. It's evident that the training process of the service processing model is entirely executed by the target service device. In other words, this decouples the engineering part (i.e., the business routing process) from the model algorithm (service processing algorithm). The engineering part has a unified basic service and routing mechanism, while the target service device only needs to focus on developing its local service processing algorithm, evolving independently. With the engineering part platformized, the continuous delivery process of the game business is also standardized and automated, shortening the time from game idea to launch. This allows for personalized adjustments to many game scenarios throughout the entire process using machine learning, improving the gaming experience.

[0096] Optionally, the above-mentioned invocation of the initial business processing model, which matches and predicts the business feature data of the sample game object and the business feature data of the sample players in the sample player set to obtain the predicted opponent of the sample game object, includes: the target business device can invoke the initial business processing model to match and predict the business feature data of the sample game object and the business feature data of the sample players in the sample player set to obtain the predicted matching degree between the sample game object and the sample players in the sample player set; the predicted matching degree is used to reflect one or more of the following: the probability of abnormal events occurring when the sample game object and the sample players play against each other, the difference in combat ability between the sample game object and the sample players, etc. Then, sample players with a predicted matching degree greater than a matching degree threshold can be selected from the sample player set, and the selected sample players are determined as the predicted opponent of the sample game object. By determining the opponent of the sample game object based on the predicted matching degree, the training accuracy of the initial business processing model is improved.

[0097] Optionally, adjusting the initial service processing model based on the predicted and labeled players to obtain a service processing model for processing the target game service includes: the target service device determining the player prediction error of the initial service processing model based on the predicted and labeled players, such as determining the player prediction error based on the matching degree between the object attribute information of the predicted players and the object attribute information of the labeled players. If the player prediction error is in a convergent state, it indicates that the player prediction error of the initial service processing model is relatively small, that is, it indicates that the player prediction error of the initial service processing model has been reduced to the minimum error value. Therefore, the initial service processing model can be determined as the service processing model for processing the target game service. If the player prediction error is not in a convergent state, it indicates that the player prediction error of the initial service processing model is relatively large, that is, it indicates that the player prediction error of the initial service processing model has been reduced to the minimum error value. Therefore, the initial service processing model can be adjusted based on the player prediction error. The adjusted initial business processing model is determined as the business processing model for handling the target game business. Adjusting the initial business processing model based on the prediction error of the opposing players helps improve the accuracy of the business processing model in identifying opposing players.

[0098] Understandably, games often involve multiple scenarios requiring business processing models. These scenarios are abstracted as game business processes, and different game business processes need to be securely isolated, such as ensuring that code and operational metrics are not visible to each other. Therefore, the business processing models corresponding to different game business processes are deployed on different business devices to achieve security isolation between them. A game business may have multiple developers, and each developer can develop and deploy their respective business processing models without affecting the others. The number of virtual game objects varies greatly, and the number of virtual game objects also varies greatly at different operational stages of a game. Therefore, the architecture must be flexible enough to scale up and down to support different request volumes. Such systems are generally distributed architectures formed by the interaction of multiple service modules. This requires that processing requests issued by the game business, after passing through multiple service modules, can be routed to the desired business processing model. The engineering part and the model algorithm should be separated, with the model part focusing only on the local algorithm. After receiving a processing request, the network management device in the business processing system starts a single transaction service. The resource management device selects a suitable business processing device, which then retrieves business feature data. The request and business feature data are then passed to the business processing model through an interface. The business processing model infers and generates business processing results, which are then sent back to the terminal device by the network management device. The statistical and offline data generated by the model are also saved and displayed by the business processing device.

[0099] S206. The target service device calls the target service processing model based on the storage address and generates scenario recommendation data based on the scenario feature data.

[0100] In steps S204 to S206, the process of calling the target service processing model based on the storage address and generating scene recommendation data according to the scene feature data via the target service device includes: when the service processing model library is inserted into the target service device as a dynamic link library, the target service device can obtain the dynamic loading function corresponding to the service processing model library, call the dynamic loading function, and obtain the storage address of the target service processing model in the service processing model library. Then, the target service processing model can be called according to the storage address, and the scene recommendation data generated according to the scene feature data can be obtained. By calling the service processing model as a dynamic link library, the convenience of obtaining the service processing model can be improved.

[0101] S207. Based on the scenario recommendation data, the target service device makes game service recommendations in the game scenario.

[0102] In this application, when a game scene meets the game service recommendation conditions, the game service platform triggers a scene recommendation request. Based on this request, it obtains a set of scene service parameters matching the game scene. According to these parameters, it filters target service devices matching the request from the service device cluster, enabling the scene recommendation request to be routed to the target device. Furthermore, it allows for the routed distribution of different scene recommendation requests to different service devices, achieving distributed processing of scene recommendation requests. Further, it can extract scene feature data matching the request from the target service device. Based on the scene recommendation data generated by processing this feature data, game service recommendations are performed within the game scene, enabling online recommendation throughout the entire game process—that is, online processing of game services for multiple game scenes. Therefore, the game service platform implements the service routing process for scene recommendation requests. This routing process determines which service device will handle the request, and the target service device handles the request. The routing and processing processes are independent of each other, achieving decoupling and facilitating efficient processing of different game services. Meanwhile, since the business diversion process and the business processing process are independent of each other, when a user has new game business requirements, only the new game business processing algorithm needs to be added to the target business device, without needing to process the business diversion process, which helps to improve the development efficiency of game business.

[0103] Understandably, this solution decouples the engineering component from the model algorithm. The engineering component has a unified basic service and a traffic distribution mechanism, while the model component only needs to focus on developing local algorithms, allowing each to evolve independently. After the engineering component is platformized, the continuous delivery process is also standardized and automated, shortening the time from game business idea to launch. This allows for personalized adjustments to a large number of game scenarios throughout the entire process using machine learning, improving the game experience.

[0104] Please see Figure 9 This is a schematic diagram of the structure of a service processing device for a game scene provided in an embodiment of this application. The aforementioned service processing device for a game scene can be a computer program (including program code) running on a network device; for example, the service processing device for the game scene is an application software. This device can be used to execute the corresponding steps in the method provided in the embodiments of this application. Figure 9 As shown, the business processing device for this game scene may include: a trigger module 911, an acquisition module 912, a filtering module 913, an extraction module 914, and a generation module 915.

[0105] The trigger module is used to trigger a scene recommendation request when the game scene meets the game business recommendation conditions;

[0106] The acquisition module is used to acquire a set of scene business parameters that match the game scene based on the scene recommendation request;

[0107] The filtering module is used to filter target business devices in the business device cluster that match the scenario recommendation request based on the scenario business parameter set.

[0108] The extraction module is used to extract scene feature data that matches the scene recommendation request through the target service device;

[0109] The generation module is used to recommend game services in the game scene based on the scene recommendation data generated by the target service device through processing the scene feature data.

[0110] Optionally, the filtering module includes a first filtering unit 91a, a first acquisition unit 92a, and a second filtering unit 93a;

[0111] The first filtering unit 91a is used to filter target resource management devices that match the scenario recommendation request from the resource management device cluster based on the scenario business parameter set.

[0112] The first acquisition unit 92a is used to acquire the business equipment cluster associated with the target resource management device;

[0113] The second filtering unit 93a is used to filter target business devices that match the scenario recommendation request from the business device cluster associated with the target resource management device based on the scenario business parameter set.

[0114] The second filtering unit 93a, based on the scenario service parameter set, filters target service devices that match the scenario recommendation request from the service device cluster associated with the target resource management device, including:

[0115] Obtain the business device relationship tree corresponding to the target resource management device; the parent node of the business device relationship tree is the target resource management device, and the child nodes of the business device relationship tree are the business devices in the business device cluster associated with the target resource management device;

[0116] Based on the node paths of the business node relationship tree, traverse the business devices in the business device relationship tree;

[0117] Based on the scenario service parameter set, target service devices that match the scenario recommendation request are selected from the service device relationship tree.

[0118] Optionally, the scenario service parameter set includes the traffic splitting keywords, traffic splitting calculation methods, and traffic splitting parameters corresponding to the game scenario; the second filtering unit 93a, based on the scenario service parameter set, filters target service devices matching the scenario recommendation request from the service device relationship tree, including:

[0119] The splitting keyword and the splitting parameter are calculated according to the splitting calculation method to obtain the splitting value of the game scene;

[0120] Obtain the numerical range corresponding to the business device in the business device relationship tree; the numerical range is generated based on the traffic splitting keyword corresponding to the game scene for which the business device has management authority and the traffic splitting parameter.

[0121] Determine the relationship between the numerical ranges corresponding to the business devices in the business device relationship tree and the split values ​​of the game scene objects;

[0122] The business devices that have an inclusion relationship with the corresponding numerical range in the business device relationship tree and the traffic distribution value of the game scene are identified as the target business devices that match the scene recommendation request.

[0123] Optionally, the scene service parameter set includes the location information of the terminal device corresponding to the virtual game object in the game scene; the second filtering unit 93a filters target service devices matching the scene recommendation request from the service device relationship tree based on the scene service parameter set, including:

[0124] Obtain the location set corresponding to the business devices in the business device relationship tree; the location set includes the location information of the terminal devices that the business devices have management authority over in the business device relationship tree;

[0125] Determine the relationship between the location information of the terminal device corresponding to the virtual game object and the set of locations corresponding to the business devices in the business device relationship tree;

[0126] The business devices that are included in the relationship between the set of locations corresponding to the business device relationship tree and the location information of the terminal device corresponding to the virtual game object are identified as the target business devices that match the scene recommendation request.

[0127] Optionally, the second filtering unit 93a, based on the scenario service parameter set, filters target service devices that match the scenario recommendation request from the service device relationship tree, including:

[0128] Based on the scenario service parameter set, select the service devices that match the scenario recommendation request from the service device relationship tree as candidate service devices;

[0129] Obtain the volume of the game services to be processed corresponding to the candidate service devices;

[0130] Candidate service devices with a service volume less than the service volume threshold are identified as target service devices that match the scenario recommendation request.

[0131] Optionally, the scene service parameter set includes the scene identifier of the game scene; the second filtering unit 93a, based on the scene service parameter set, filters target service devices matching the scene recommendation request from the service device relationship tree, including:

[0132] Based on the scene identifier of the game scene, determine the target business processing model for handling the scene recommendation request;

[0133] The business devices that have the business processing model in the business device relationship tree are selected as the target business devices that match the scenario recommendation request.

[0134] Optionally, the second filtering unit 93a determines a target business processing model for processing the scene recommendation request based on the scene identifier of the game scene, including:

[0135] Obtain the business processing model library used to process the scenario recommendation request, and the business processing logs of the business processing models in the business processing model library within a historical time period;

[0136] Determine the business processing performance parameters of the business processing models in the business processing model library based on the business processing logs;

[0137] Based on the business processing performance parameters, a target business processing model for processing the scenario recommendation request is determined from the business processing model library.

[0138] Optionally, the generation module 915 includes a second acquisition unit 94b, a calling unit 95b, and a generation unit 96b:

[0139] The second acquisition unit 94b is used to acquire the dynamic loading function corresponding to the business processing model library;

[0140] Calling unit 95b is used to call the dynamic loading function through the target service device to obtain the storage address of the target service processing model in the service processing model library;

[0141] The generation unit 96b is used to call the target service processing model based on the storage address through the target service device, and to recommend game services in the game scene according to the scene recommendation data.

[0142] Optionally, the scene recommendation request is used to instruct the matching of game players to virtual game objects in the game scene; the generation unit 96b calls the target service processing model based on the storage address through the target service device, and generates scene recommendation data according to the scene feature data, including:

[0143] The target service device obtains the service feature data of the candidate players in the candidate player set regarding the game scene, as well as the service feature data of the virtual game object, from the scene feature data;

[0144] The target business processing model is invoked through the target business device to match and identify the business feature data of the virtual game object and the business feature data of the candidate players in the candidate player set, thereby obtaining the matching degree between the virtual game object and the candidate players in the candidate player set.

[0145] The target service device selects candidate players from the candidate player set whose matching degree is greater than the matching degree threshold, and uses them as the players in the virtual game object; scene recommendation data is generated based on the players in the virtual game object.

[0146] Optionally, the generation unit 96b calls the target service processing model through the target service device to match and identify the service feature data of the virtual game object and the service feature data of the candidate players in the candidate player set, to obtain the matching degree between the virtual game object and the candidate players in the candidate player set, including:

[0147] The abnormal event recognition layer of the target business processing model is invoked to perform abnormal event recognition on the business feature data of the virtual game object and the business feature data of the candidate players in the candidate player set, and to obtain the abnormal event probability; the abnormal event probability is used to reflect the probability of an abnormal event occurring when the virtual game object and the candidate players in the candidate player set are in battle.

[0148] The capability recognition layer of the target business processing model is invoked to identify the combat capabilities of the virtual game object based on the business characteristic data of the virtual game object.

[0149] Identify the combat capabilities of candidate players in the candidate player set based on their business characteristic data;

[0150] The matching degree between the virtual game object and the candidate players in the candidate player set is determined based on the probability of the abnormal event, the combat ability of the virtual game object, and the combat ability of the candidate players in the candidate player set.

[0151] Optionally, the generation unit 96b determines the matching degree between the virtual game object and the candidate players in the candidate player set based on the probability of the abnormal event, the combat ability of the virtual game object, and the combat ability of the candidate players in the candidate player set, including:

[0152] Determine the difference in combat ability between the virtual game object and the candidate players in the candidate player set;

[0153] The difference in combat ability and the probability of abnormal events are weighted and summed to obtain the combat difference value between the virtual game object and the candidate players in the candidate player set.

[0154] The matching degree between the virtual game object and the candidate players in the candidate player set is determined based on the battle difference value; there is a negative correlation between the battle difference value and the matching degree between the virtual game object and the candidate players in the candidate player set.

[0155] Optionally, the acquisition module is further configured to acquire business feature data of the sample game object regarding the scene recommendation request, the labeled players of the sample game object, and business feature data of the sample players in the sample player set regarding the scene recommendation request; the scene recommendation request is used to instruct the allocation of players to the sample game object.

[0156] The calling module is used to call the initial business processing model, and perform matching and prediction based on the business feature data of the sample game object and the business feature data of the sample players in the sample player set to obtain the predicted players of the sample game object; the predicted players belong to the sample player set.

[0157] The adjustment module is used to adjust the initial business processing model based on the predicted players and the labeled players to obtain a target business processing model for processing the scenario recommendation business.

[0158] An add module is used to add the processing model for the target business to the business processing model library.

[0159] Optionally, the calling module invokes the initial business processing model to perform matching and prediction based on the business feature data of the sample game object and the business feature data of the sample players in the sample player set, to obtain the predicted opposing players of the sample game object, including:

[0160] The initial business processing model is invoked to match and predict the business feature data of the sample game object and the business feature data of the sample players in the sample player set, so as to obtain the predicted matching degree between the sample game object and the sample players in the sample player set.

[0161] Select sample players from the set of sample players whose predicted matching degree is greater than the matching degree threshold;

[0162] The selected sample players are determined as the predicted opponents for the sample game object.

[0163] Optionally, the adjustment module adjusts the initial business processing model based on the predicted and labeled players to obtain a business processing model for processing the target game business, including:

[0164] Based on the predicted players and the labeled players, the player prediction error of the initial business processing model is determined.

[0165] If the prediction error of the player in the battle is not in a convergent state, the initial business processing model shall be adjusted according to the prediction error of the player in the battle.

[0166] The adjusted initial business processing model is determined as the business processing model used to process the target game business.

[0167] Optionally, when the game scene meets the game service recommendation conditions, the triggering unit triggers a scene recommendation request, including:

[0168] If the game difficulty of the game scene meets the adjustment conditions, then the game scene is determined to meet the game service recommendation conditions, and a scene recommendation request is generated to indicate the adjustment of the game difficulty level of the game scene;

[0169] If the game scene meets the game item recommendation criteria, then the game scene is determined to meet the game service recommendation criteria, and a scene recommendation request is generated to instruct the game scene to recommend game items.

[0170] If an instruction to match game players for the game scene is received, it is determined that the game scene meets the game service recommendation conditions, and a scene recommendation request is generated to instruct the matching of game players to virtual game objects in the game scene.

[0171] According to one embodiment of this application, Figure 3 The steps involved in the business processing method of the game scenario shown can be derived from... Figure 9 The various modules in the business processing device of the game scene shown are executed. For example, step S101 shown in Figure 2 can be performed by... Figure 9The trigger module 911 in the middle is used to execute, Figure 3 Step S102 shown can be performed by Figure 9 The acquisition module 912 in the middle is used to execute; Figure 3 Step S103 shown can be performed by Figure 9 The filtering module 913 in the middle is used to perform the operation. Figure 3 Step S104 shown can be derived from Figure 9 The extraction module 914 in the middle is used to execute; Figure 3 Step S105 shown can be performed by Figure 9 The generation module 915 in the middle is used to execute.

[0172] According to one embodiment of this application, Figure 9 The modules in the service processing device for the game scene shown can be individually or entirely merged into one or more units, or some of these units can be further divided into at least two functionally smaller sub-units to achieve the same operation without affecting the technical effects of the embodiments of this application. The above modules are based on logical functional division. In practical applications, the function of one module can be implemented by at least two units, or the function of at least two modules can be implemented by one unit. In other embodiments of this application, the service processing device for the game scene may also include other units. In practical applications, these functions can also be implemented with the assistance of other units, and can be implemented collaboratively by at least two units.

[0173] According to one embodiment of this application, the following can be executed by running on a general-purpose computer device, such as a computer, which includes processing components and storage components such as a central processing unit (CPU), random access memory (RAM), and read-only memory (ROM). Figure 3 The computer program (including program code) for each step involved in the corresponding method shown, to construct such... Figure 9 The present invention describes a business processing apparatus for a game scene and a business processing method for implementing the game scene in the embodiments of this application. The computer program described above may be recorded on, for example, a computer-readable recording medium, loaded onto the aforementioned computing device via the computer-readable recording medium, and run therein.

[0174] In this application, when a game scene meets the game service recommendation conditions, the game service platform triggers a scene recommendation request. Based on this request, it obtains a set of scene service parameters matching the game scene. According to these parameters, it filters target service devices matching the request from the service device cluster, enabling the scene recommendation request to be routed to the target device. Furthermore, it allows for the routed distribution of different scene recommendation requests to different service devices, achieving distributed processing of scene recommendation requests. Further, it can extract scene feature data matching the request from the target service device. Based on the scene recommendation data generated by processing this feature data, game service recommendations are performed within the game scene, enabling online recommendation throughout the entire game process—that is, online processing of game services for multiple game scenes. Therefore, the game service platform implements the service routing process for scene recommendation requests. This routing process determines which service device will handle the request, and the target service device handles the request. The routing and processing processes are independent of each other, achieving decoupling and facilitating efficient processing of different game services. Meanwhile, since the business diversion process and the business processing process are independent of each other, when a user has new game business requirements, only the new game business processing algorithm needs to be added to the target business device, without needing to process the business diversion process, which helps to improve the development efficiency of game business.

[0175] Please see Figure 10 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Figure 10 As shown, the computer device 1000 may include a processor 1001, a network interface 1004, and a memory 1005. Furthermore, the computer device 1000 may also include a user interface 1003 and at least one communication bus 1002. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen and a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be high-speed RAM or non-volatile memory, such as at least one disk storage device. Optionally, the memory 1005 may also be at least one storage device located remotely from the processor 1001. Figure 10 As shown, the memory 1005, which is a computer-readable storage medium, may include an operating system, a network communication module, a user interface module, and a device control application.

[0176] exist Figure 10 In the computer device 1000 shown, the network interface 1004 provides network communication functionality; the user interface 1003 is mainly used to provide an input interface; and the processor 1001 can be used to call the device control application stored in the memory 1005 to achieve:

[0177] When a game scene meets the game business recommendation criteria, a scene recommendation request is triggered.

[0178] Based on the scenario recommendation request, obtain the scenario business parameter set that matches the game scenario;

[0179] Based on the scenario service parameter set, target service devices that match the scenario recommendation request are selected from the service device cluster;

[0180] The scene feature data matching the scene recommendation request is extracted from the target service device;

[0181] Based on the scene recommendation data generated by the target service device processing the scene feature data, game service recommendations are made in the game scene.

[0182] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to filter target service devices matching the scenario recommendation request in the service device cluster based on the scenario service parameter set, including:

[0183] Based on the scenario service parameter set, target resource management devices that match the scenario recommendation request are selected from the resource management device cluster;

[0184] Obtain the business device cluster associated with the target resource management device;

[0185] Based on the scenario service parameter set, target service devices that match the scenario recommendation request are selected from the service device cluster associated with the target resource management device.

[0186] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to filter target service devices matching the scenario recommendation request from the service device cluster associated with the target resource management device based on the scenario service parameter set, including:

[0187] Obtain the business device relationship tree corresponding to the target resource management device; the parent node of the business device relationship tree is the target resource management device, and the child nodes of the business device relationship tree are the business devices in the business device cluster associated with the target resource management device;

[0188] Based on the node paths of the business node relationship tree, traverse the business devices in the business device relationship tree;

[0189] Based on the scenario service parameter set, target service devices that match the scenario recommendation request are selected from the service device relationship tree.

[0190] Optionally, the scene service parameter set includes the traffic splitting keyword, traffic splitting operation method, and traffic splitting parameters corresponding to the game scene; the processor 1001 can be used to call the device control application stored in the memory 1005 to filter target service devices matching the scene recommendation request from the service device relationship tree based on the scene service parameter set, including:

[0191] The splitting keyword and the splitting parameter are calculated according to the splitting calculation method to obtain the splitting value of the game scene;

[0192] Obtain the numerical range corresponding to the business device in the business device relationship tree; the numerical range is generated based on the traffic splitting keyword corresponding to the game scene for which the business device has management authority and the traffic splitting parameter.

[0193] Determine the relationship between the numerical ranges corresponding to the business devices in the business device relationship tree and the split values ​​of the game scene objects;

[0194] The business devices that have an inclusion relationship with the corresponding numerical range in the business device relationship tree and the traffic distribution value of the game scene are identified as the target business devices that match the scene recommendation request.

[0195] Optionally, the scene service parameter set includes the location information of the terminal device corresponding to the virtual game object in the game scene;

[0196] Processor 1001 can be used to call the device control application stored in memory 1005 to filter target service devices matching the scenario recommendation request from the service device relationship tree based on the scenario service parameter set, including:

[0197] Obtain the location set corresponding to the business devices in the business device relationship tree; the location set includes the location information of the terminal devices that the business devices have management authority over in the business device relationship tree;

[0198] Determine the relationship between the location information of the terminal device corresponding to the virtual game object and the set of locations corresponding to the business devices in the business device relationship tree;

[0199] The target business device that matches the scenario recommendation request is determined by identifying the business devices that have an inclusion relationship with the location set corresponding to the business device relationship tree and the location information of the terminal device corresponding to the virtual game object.

[0200] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to filter target service devices matching the scenario recommendation request from the service device relationship tree based on the scenario service parameter set, including:

[0201] Based on the scenario service parameter set, select the service devices that match the scenario recommendation request from the service device relationship tree as candidate service devices;

[0202] Obtain the volume of the game services to be processed corresponding to the candidate service devices;

[0203] Candidate service devices with a service volume less than the service volume threshold are identified as target service devices that match the scenario recommendation request.

[0204] Optionally, the scene service parameter set includes the scene identifier of the game scene; the processor 1001 can be used to call the device control application stored in the memory 1005 to filter target service devices matching the scene recommendation request from the service device relationship tree based on the scene service parameter set, including:

[0205] Based on the scene identifier of the game scene, determine the target business processing model for handling the scene recommendation request;

[0206] The business devices that have the business processing model in the business device relationship tree are selected as the target business devices that match the scenario recommendation request.

[0207] Optionally, the processor 1001 can be used to invoke a device control application stored in the memory 1005 to determine a target business processing model for processing the scene recommendation request based on the scene identifier of the game scene, including:

[0208] Obtain the business processing model library used to process the scenario recommendation request, and the business processing logs of the business processing models in the business processing model library within a historical time period;

[0209] Determine the business processing performance parameters of the business processing models in the business processing model library based on the business processing logs;

[0210] Based on the business processing performance parameters, a target business processing model for processing the scenario recommendation request is determined from the business processing model library.

[0211] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to perform game service recommendations in the game scene based on the scene feature data generated by the target service device processing the scene feature data, including:

[0212] The dynamic loading function corresponding to the business processing model library is obtained through the target business device;

[0213] The target service device calls the dynamic loading function to obtain the storage address of the target service processing model in the service processing model library;

[0214] The target service device invokes the target service processing model based on the storage address, generates scene recommendation data according to the scene feature data, and recommends game services in the game scene based on the scene recommendation data.

[0215] Optionally, the scene recommendation request is used to instruct the matching of game players to virtual game objects in the game scene; the processor 1001 can be used to call the device control application stored in the memory 1005 to realize the scene recommendation data generated by calling the target service processing model based on the storage address through the target service device, including:

[0216] The target service device obtains the service feature data of the candidate players in the candidate player set regarding the game scene, as well as the service feature data of the virtual game object, from the scene feature data;

[0217] The target business processing model is invoked through the target business device to match and identify the business feature data of the virtual game object and the business feature data of the candidate players in the candidate player set, thereby obtaining the matching degree between the virtual game object and the candidate players in the candidate player set.

[0218] The target service device selects candidate players from the candidate player set whose matching degree is greater than the matching degree threshold, and uses them as the players in the virtual game object; scene recommendation data is generated based on the players in the virtual game object.

[0219] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to invoke the target business processing model, match and identify the business feature data of the virtual game object and the business feature data of the candidate players in the candidate player set, and obtain the matching degree between the virtual game object and the candidate players in the candidate player set, including:

[0220] The abnormal event recognition layer of the target business processing model is invoked to perform abnormal event recognition on the business feature data of the virtual game object and the business feature data of the candidate players in the candidate player set, and to obtain the abnormal event probability; the abnormal event probability is used to reflect the probability of an abnormal event occurring when the virtual game object and the candidate players in the candidate player set are in battle.

[0221] The capability recognition layer of the target business processing model is invoked to identify the combat capabilities of the virtual game object based on the business characteristic data of the virtual game object.

[0222] Identify the combat capabilities of candidate players in the candidate player set based on their business characteristic data;

[0223] The matching degree between the virtual game object and the candidate players in the candidate player set is determined based on the probability of the abnormal event, the combat ability of the virtual game object, and the combat ability of the candidate players in the candidate player set.

[0224] Optionally, the processor 1001 can be used to call a device control application stored in the memory 1005 to determine the matching degree between the virtual game object and the candidate players in the candidate player set based on the probability of the abnormal event, the combat ability of the virtual game object, and the combat ability of the candidate players in the candidate player set, including:

[0225] The difference in combat ability between the virtual game object and the candidate players in the candidate player set is determined by the target service device.

[0226] The difference in combat capability and the probability of abnormal events are weighted and summed by the target service device to obtain the combat difference value between the virtual game object and the candidate players in the candidate player set.

[0227] The target service device determines the matching degree between the virtual game object and the candidate players in the candidate player set based on the battle difference value; there is a negative correlation between the battle difference value and the matching degree between the virtual game object and the candidate players in the candidate player set.

[0228] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to: acquire business feature data of sample game objects regarding the scene recommendation request, the labeled players of the sample game objects, and business feature data of sample players in the sample player set regarding the scene recommendation request; the scene recommendation request is used to instruct the matching of game players to virtual game objects in the game scene;

[0229] The initial business processing model is invoked, and matching and prediction are performed based on the business feature data of the sample game object and the business feature data of the sample players in the sample player set to obtain the predicted opposing players of the sample game object; the predicted opposing players belong to the sample player set.

[0230] Based on the predicted players and the labeled players, the initial business processing model is adjusted to obtain a target business processing model for processing the scenario recommendation request.

[0231] Add the target business processing model to the business processing model library.

[0232] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to: call the initial business processing model, and perform matching and prediction based on the business feature data of the sample game object and the business feature data of the sample players in the sample player set to obtain the predicted opponent of the sample game object, including:

[0233] The initial business processing model is invoked to match and predict the business feature data of the sample game object and the business feature data of the sample players in the sample player set, so as to obtain the predicted matching degree between the sample game object and the sample players in the sample player set.

[0234] Select sample players from the set of sample players whose predicted matching degree is greater than the matching degree threshold;

[0235] The selected sample players are determined as the predicted opponents for the sample game object.

[0236] Optionally, the processor 1001 can be used to call the device control application stored in the memory 1005 to trigger a scene recommendation request when the game scene meets the game business recommendation conditions, including:

[0237] If the game difficulty of the game scene meets the adjustment conditions, then the game scene is determined to meet the game service recommendation conditions, and a scene recommendation request is generated to indicate the adjustment of the game difficulty level of the game scene;

[0238] If the game scene meets the game item recommendation criteria, then the game scene is determined to meet the game service recommendation criteria, and a scene recommendation request is generated to instruct the game scene to recommend game items.

[0239] If an instruction to match game players for the game scene is received, it is determined that the game scene meets the game service recommendation conditions, and a scene recommendation request is generated to instruct the matching of game players to virtual game objects in the game scene.

[0240] In this application, when a game scene meets the game service recommendation conditions, the game service platform triggers a scene recommendation request. Based on this request, it obtains a set of scene service parameters matching the game scene. According to these parameters, it filters target service devices matching the request from the service device cluster. This allows the scene recommendation request to be routed to the target device, and different scene recommendation requests to different devices, achieving distributed processing. Furthermore, scene feature data matching the request can be extracted from the target device, and game service recommendations can be performed within the game scene, enabling online recommendations throughout the entire game process—that is, online processing of game services for multiple game scenes. Therefore, the game service platform implements the service routing process for scene recommendation requests. This routing process determines which service device handles the request, and the target device implements the service processing. The routing and processing processes are independent, decoupling them and facilitating efficient processing of different game services. Meanwhile, since the business diversion process and the business processing process are independent of each other, when a user has new game business requirements, only the new game business processing algorithm needs to be added to the target business device, without needing to process the business diversion process, which helps to improve the development efficiency of game business.

[0241] It should be understood that the computer device 1000 described in the embodiments of this application can execute the foregoing text. Figure 3 or Figure 5 The description of the business processing method for the game scene in the corresponding embodiment can also be executed as described above. Figure 9 and Figure 10 The description of the business processing device for the game scene in the corresponding embodiments will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated.

[0242] Furthermore, it should be noted that this application embodiment also provides a computer-readable storage medium, which stores a computer program executed by the business processing device for the game scene mentioned above. The computer program includes program instructions, and when the processor executes the program instructions, it can execute the aforementioned... Figure 3 and the preceding text Figure 5The description of the business processing method for the above game scenario in the corresponding embodiments will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the computer-readable storage medium embodiments related to this application, please refer to the description of the method embodiments of this application.

[0243] As an example, the above program instructions can be deployed and executed on a computer device, or deployed and executed on at least two computer devices in one location, or executed on at least two computer devices distributed in at least two locations and interconnected by a communication network. At least two computer devices distributed in at least two locations and interconnected by a communication network can form a blockchain network.

[0244] The aforementioned computer-readable storage medium can be a business processing device for the game scene provided in any of the foregoing embodiments, or a central storage unit of the aforementioned computer device, such as a hard disk or central storage of the computer device. The computer-readable storage medium can also be an external storage device of the computer device, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device. Furthermore, the computer-readable storage medium can include both the central storage unit and external storage devices of the computer device. The computer-readable storage medium is used to store the computer program and other programs and data required by the computer device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

[0245] The terms "first," "second," etc., in the specification, claims, and drawings of this application are used to distinguish content in different media, rather than to describe a specific order. Furthermore, the term "comprising," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or modules, but may optionally include steps or modules not listed, or may optionally include other step units inherent to these processes, methods, apparatuses, products, or devices.

[0246] This application also provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the foregoing description. Figure 3 and Figure 5The description of the business processing method for the above game scenario in the corresponding embodiments will not be repeated here. Furthermore, the beneficial effects of using the same method will also not be repeated. For technical details not disclosed in the embodiments of the computer program product involved in this application, please refer to the description of the method embodiments of this application.

[0247] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0248] The methods and related apparatus provided in this application are described with reference to the method flowcharts and / or structural diagrams provided in this application. Specifically, each block of the method flowchart and / or structural diagram, as well as combinations of blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable network-connected device to create a machine, such that the instructions, which execute via the processor of the computer or other programmable network-connected device, generate instructions for implementing the process. Figure 1 A schematic diagram of one or more processes and / or structures. Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable network-connected device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 A schematic diagram of one or more processes and / or structures. Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable network-connected device to cause a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable device for implementing the process. Figure 1 A process or multiple processes and / or structures illustrate the steps of the functions specified in one or more boxes.

[0249] The above-disclosed embodiments are merely preferred embodiments of this application and should not be construed as limiting the scope of this application. Therefore, any equivalent variations made in accordance with the claims of this application shall still fall within the scope of this application.

Claims

1. A service processing method of a game scene, characterized in that, The method comprises the following steps: When the game scene meets the game service recommendation condition, a scene recommendation request is triggered; Based on the scene recommendation request, a set of scene service parameters matching the game scene is obtained; Based on the set of scene service parameters, a target service device matching the scene recommendation request is screened in a service device cluster; Through the target service device, scene feature data matching the scene recommendation request is extracted; Through the target service device, a dynamic loading function corresponding to a service processing model library is obtained; the service processing model library is used to process the scene recommendation request; Through the target service device, the storage address of a target service processing model in the service processing model library is obtained by calling the dynamic loading function; the target service processing model is determined based on the scene identifier of the game scene; Through the target service device, scene recommendation data is generated based on the scene feature data by calling the target service processing model based on the storage address; game service recommendation is performed in the game scene based on the scene recommendation data.

2. The method of claim 1, wherein, Based on the set of scene service parameters, a target service device matching the scene recommendation request is screened in a service device cluster, which comprises the following steps: Based on the set of scene service parameters, a target resource management device matching the scene recommendation request is screened from a resource management device cluster; A service device cluster associated with the target resource management device is obtained; Based on the set of scene service parameters, a target service device matching the scene recommendation request is screened from the service device cluster associated with the target resource management device.

3. The method of claim 2, wherein, Based on the set of scene service parameters, a target service device matching the scene recommendation request is screened from the service device cluster associated with the target resource management device, which comprises the following steps: A business device relationship tree corresponding to the target resource management device is obtained; the parent node of the business device relationship tree is the target resource management device, and the child node of the business device relationship tree is the business device in the business device cluster associated with the target resource management device; According to the node path of the business node relationship tree, the business devices in the business device relationship tree are traversed; Based on the set of scene service parameters, a target service device matching the scene recommendation request is screened from the business device relationship tree.

4. The method of claim 3, wherein, The set of scene service parameters comprises a shunting keyword, a shunting operation method, and a shunting parameter corresponding to the game scene; Based on the set of scene service parameters, a target service device matching the scene recommendation request is screened from the business device relationship tree, which comprises the following steps: According to the shunting operation method, the shunting keyword and the shunting parameter are operated to obtain the shunting value of the game scene; The value interval corresponding to the business device in the business device relationship tree is obtained; the value interval is generated according to the shunting keyword and the shunting parameter corresponding to the game scene managed by the business device; The relationship between the value interval corresponding to the business device in the business device relationship tree and the shunting value of the game scene object is determined; The business device corresponding to the relationship between the corresponding numerical interval in the business device relationship tree and the shunting numerical value of the game scene is determined as a target business device matching the business recommendation request.

5. The method of claim 3, wherein, The scene business parameter set includes position information of a terminal device corresponding to a virtual game object in the game scene; The filtering of the target business device matching the scene recommendation request from the business device relationship tree based on the scene business parameter set includes: Obtaining a position set corresponding to a business device in the business device relationship tree; The position set includes position information of a terminal device to which the business device in the business device relationship tree has management authority; Determine the relationship between the position information of the terminal device corresponding to the virtual game object and the position set corresponding to the business device in the business device relationship tree; The business device corresponding to the relationship between the corresponding position set in the business device relationship tree and the position information of the terminal device corresponding to the virtual game object is determined as the target business device matching the scene recommendation request.

6. The method of claim 3, wherein, The filtering of the target business device matching the scene recommendation request from the business device relationship tree based on the scene business parameter set includes: Filtering the business device matching the business recommendation request from the business device relationship tree based on the scene business parameter set as a candidate business device; Obtaining the business volume corresponding to the to-be-processed game business of the candidate business device; The candidate business device with a business volume less than a business volume threshold is determined as the target business device matching the scene recommendation request.

7. The method of claim 3, wherein, The scene business parameter set includes a scene identifier of the game scene; The filtering of the target business device matching the scene recommendation request from the business device relationship tree based on the scene business parameter set includes: Determine the target business processing model used to process the scene recommendation request according to the scene identifier of the game scene; The business device in the business device relationship tree with the business processing model is determined as the target business device matching the scene recommendation request.

8. The method of claim 7, wherein, The determination of the target business processing model used to process the scene recommendation request according to the scene identifier of the game scene includes: Obtaining a business processing model library used to process the scene recommendation request, and a business processing log of a business processing model in the business processing model library in a historical time period; Determine the business processing performance parameter of the business processing model in the business processing model library according to the business processing log; Determine the target business processing model used to process the scene recommendation request from the business processing model library according to the business processing performance parameter.

9. The method of claim 1, wherein, The scene recommendation request is used to indicate matching game players to virtual game objects in the game scene; The generation of scene recommendation data according to the scene feature data by the target business device based on the storage address calling the target business processing model includes: obtaining, by the target business device, from the scene feature data, business feature data of a candidate player in a candidate player set with respect to the game scene and business feature data of the virtual game object; calling, by the target business device, the target business processing model to perform matching identification on the business feature data of the virtual game object and the business feature data of the candidate player in the candidate player set, to obtain a matching degree between the virtual game object and the candidate player in the candidate player set; selecting, by the target business device, from the candidate player set, a candidate player corresponding to a matching degree greater than a matching degree threshold, as a battle player of the virtual game object; generating scene recommendation data according to the battle player of the virtual game object.

10. The method of claim 9, wherein, The calling, by the target business device, of the target business processing model to perform matching identification on the business feature data of the virtual game object and the business feature data of the candidate player in the candidate player set, to obtain a matching degree between the virtual game object and the candidate player in the candidate player set, includes: calling, by the target business device, an abnormal event identification layer of the target business processing model to perform abnormal event identification on the business feature data of the virtual game object and the business feature data of the candidate player in the candidate player set, to obtain an abnormal event probability; the abnormal event probability is used to reflect a probability of occurrence of an abnormal event when the virtual game object and the candidate player in the candidate player set battle; calling, by the target business device, a capability identification layer of the target business processing model to identify a battle capability of the virtual game object according to the business feature data of the virtual game object; identifying, by the target business device, a battle capability of the candidate player in the candidate player set according to the business feature data of the candidate player in the candidate player set; determining, by the target business device, a matching degree between the virtual game object and the candidate player in the candidate player set according to the abnormal event probability, the battle capability of the virtual game object, and the battle capability of the candidate player in the candidate player set.

11. The method of claim 10, wherein, The determining, by the target business device, of a matching degree between the virtual game object and the candidate player in the candidate player set according to the abnormal event probability, the battle capability of the virtual game object, and the battle capability of the candidate player in the candidate player set, includes: determining, by the target business device, a battle capability difference between the battle capability of the virtual game object and the battle capability of the candidate player in the candidate player set; performing, by the target business device, weighted summation on the battle capability difference and the abnormal event probability, to obtain a battle difference value between the virtual game object and the candidate player in the candidate player set; determining, by the target business device, a matching degree between the virtual game object and the candidate player in the candidate player set according to the battle difference value; the battle difference value and the matching degree between the virtual game object and the candidate player in the candidate player set have a negative correlation.

12. The method of claim 8, wherein, The method further comprises: obtaining, by the target service device, service feature data of a sample game object about the scene recommendation request, labeled battle players of the sample game object, and service feature data of sample players in a sample player set about the scene recommendation request; the scene recommendation request is used to indicate matching game players to virtual game objects in the game scene; calling an initial service processing model, performing matching prediction according to the service feature data of the sample game object and the service feature data of the sample players in the sample player set, to obtain predicted battle players of the sample game object; the predicted battle players belong to the sample player set; adjusting the initial service processing model according to the predicted battle players and the labeled battle players, to obtain a target service processing model used to process the scene recommendation request; adding the target service processing model to the service processing model library.

13. The method of claim 12, wherein, The calling of the initial service processing model, the matching prediction according to the service feature data of the sample game object and the service feature data of the sample players in the sample player set, and the obtaining of the predicted battle players of the sample game object comprise: calling the initial service processing model, performing matching prediction on the service feature data of the sample game object and the service feature data of the sample players in the sample player set, to obtain a predicted matching degree between the sample game object and the sample players in the sample player set; selecting, from the sample player set, sample players corresponding to the predicted matching degree greater than a matching degree threshold; determining the selected sample players as the predicted battle players of the sample game object.

14. The method of claim 1, wherein, The triggering of the scene recommendation request when the game scene meets the game service recommendation condition comprises: if the game difficulty of the game scene meets an adjustment condition, determining that the game scene meets the game service recommendation condition, and generating a scene recommendation request used to indicate adjusting the game difficulty level of the game scene; if the game scene meets a game prop recommendation condition, determining that the game scene meets the game service recommendation condition, and generating a scene recommendation request used to indicate recommending game props to the game scene; if an instruction of matching game players to the game scene is received, determining that the game scene meets the game service recommendation condition, and generating a scene recommendation request used to indicate matching game players to virtual game objects in the game scene.

15. A service processing apparatus of a game scene, characterized by comprising: comprise: a triggering module configured to trigger a scene recommendation request when a game scene meets a game service recommendation condition; an obtaining module configured to obtain a set of scene service parameters matched with the game scene based on the scene recommendation request; a screening module configured to screen, based on the set of scene service parameters, a target service device matched with the scene recommendation request from a service device cluster; an extracting module configured to extract, by the target service device, scene feature data matched with the scene recommendation request; a generating module comprises a second obtaining unit, a calling unit, and a generating unit: A second acquisition unit is configured to acquire, by the target service device, a dynamic loading function corresponding to a service processing model library; the service processing model library is configured to process the scene recommendation request; A calling unit is configured to call, by the target service device, the dynamic loading function to acquire a storage address of a target service processing model in the service processing model library; the target service processing model is determined based on a scene identifier of the game scene; A generating unit is configured to call, by the target service device, the target service processing model based on the storage address, generate scene recommendation data according to the scene characteristic data, and perform game service recommendation in the game scene according to the scene recommendation data. 16.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-15. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 14.

17. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 14.

18. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method in any one of claims 1 to 14.

Images