Game information prompting method and system and computer program product

Abstract

The invention provides a game information prompting method and system and a computer program product, and relates to the technical field of virtual games. The game information prompting method comprises the steps that a target game task and the current position of a first virtual character are determined; determining a path-finding path from the current position to a specified position through a path planning model, and displaying the path-finding path in a graphical user interface; generating task guidance information in the graphical user interface according to real-time map data corresponding to the first game map in the process that the first virtual character moves to the specified position along the way-finding path, so as to reduce the task difficulty of completing the target game task by the first virtual character; wherein the real-time map data comprises at least one of the following data: map resource distribution, a role position of the second virtual role, first role data corresponding to the first virtual role, second role data corresponding to the second virtual role, a position of a virtual shelter and a safe area range.

Description

Technical Field

[0001] The present invention relates to the field of virtual game technology, and in particular to a game information prompting method, system and computer program product. Background Art

[0002] With the rapid development of the electronic game industry, the types and gameplay of virtual games are becoming increasingly diverse, and the choices and challenges faced by players (or users) in virtual games are becoming more complex. In some virtual games, players need to familiarize themselves with the game environment by increasing the number of games played, such as terrain, routes, resources, or enemy distribution, so as to accumulate player experience and then formulate game strategies based on the accumulated experience. This is too high a threshold for ordinary players, especially novice players, and the gaming experience is poor. Therefore, how to help players better familiarize themselves with the gaming environment, lower the gaming threshold, and improve the gaming experience of players is an urgent problem to be solved.

[0003] In view of this, some embodiments of this specification provide a game information prompt method, system and computer program product, aiming to lower the gaming threshold and enhance the gaming experience of players. Summary of the invention

[0004] One or more embodiments of the present specification provide a method for prompting game information, wherein a graphical user interface is provided through a terminal device, wherein at least part of a game match scene is displayed in the graphical user interface, wherein the game match scene includes a first virtual character participating in the game match and a second virtual character in a different game camp from the first virtual character, wherein the first virtual character is controlled through the terminal device, and the game match corresponds to a first game map, and the method comprises: determining a target game task and a current position of the first virtual character, wherein the target game task is used to instruct the first virtual character to complete a specified game behavior at a specified position in the game match scene; determining a pathfinding path from the current position to the specified position through a path planning model, and displaying the path in the graphical user interface, wherein the path The planning model is trained by first historical game match data corresponding to the first game map; in the process of the first virtual character moving along the pathfinding path to the designated position, task guidance information is generated in the graphical user interface according to the real-time map data corresponding to the first game map, so as to reduce the task difficulty of the first virtual character to complete the target game task; wherein the real-time map data includes at least one of the following data: map resource distribution, character position of the second virtual character, first character data corresponding to the first virtual character, second character data corresponding to the second virtual character, the position of the virtual bunker, and the range of the safe area; the task guidance information includes at least one of the following data: game behavior recommendation information, moving route prompt information, warning event prompt information, and combat success rate prediction information.

[0005] One or more embodiments of the present specification also provide a game information prompt system, which provides a graphical user interface through a terminal device, and displays at least part of a game game scene in the graphical user interface, wherein the game game scene includes a first virtual character participating in the game game and a second virtual character in a different game camp from the first virtual character, wherein the first virtual character is controlled through the terminal device, and the game game corresponds to a first game map, and the system includes: a determination module, which is used to determine a target game task and a current position of the first virtual character, wherein the target game task is used to instruct the first virtual character to complete a specified game behavior at a specified position in the game game scene; and a path planning module, which is used to determine a path finding path from the current position to the specified position through a path planning model, and display it in the graphical user interface, wherein In the method, the path planning model is trained by first historical game match data corresponding to the first game map; the information guidance module is used to generate task guidance information in a graphical user interface according to real-time map data corresponding to the first game map during the process of the first virtual character moving along the path-finding path to the specified position, so as to reduce the task difficulty of the first virtual character to complete the target game task; wherein the real-time map data includes at least one of the following data: map resource distribution, character position of the second virtual character, first character data corresponding to the first virtual character, second character data corresponding to the second virtual character, the position of the virtual bunker, and the range of the safe area; the task guidance information includes at least one of the following data: game behavior recommendation information, moving route prompt information, warning event prompt information, and combat success rate prediction information.

[0006] One or more embodiments of the present specification also provide a computer program product, including a computer program, which, when at least a portion of the computer program is executed by a processor, can implement the game information prompting method of some embodiments of the present specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] This specification will be further explained by way of exemplary embodiments, which will be described in detail by way of the accompanying drawings. The same numbers in the drawings represent the same structures or steps.

[0008] Figure 1 It is a schematic diagram of a virtual game system according to some embodiments of this specification.

[0009] Figure 2 This is an exemplary flow chart of a game information prompt method shown in some embodiments of this specification.

[0010] Figure 3 This is an exemplary flow chart of a method for determining gaming behavior recommendation information according to some embodiments of this specification.

[0011] Figure 4It is a schematic diagram of a graphical user interface shown in some embodiments of this specification.

[0012] Figure 5 This is an exemplary flowchart of a method for determining mobile route prompt information according to some embodiments of this specification.

[0013] Figure 6 This is another schematic diagram of a graphical user interface shown in some embodiments of this specification.

[0014] Figure 7 This is an exemplary flowchart of another method for determining moving route prompt information according to some embodiments of this specification.

[0015] Figure 8 It is an exemplary flow chart of a method for determining warning event prompt information according to some embodiments of this specification.

[0016] Fig. 9 This is another schematic diagram of a graphical user interface shown in some embodiments of this specification.

[0017] Fig.10 It is an exemplary block diagram of a game information prompt system according to some embodiments of this specification. DETAILED DESCRIPTION

[0018] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the embodiments will be described in detail below with reference to the accompanying drawings. Obviously, the contents described below are some examples or embodiments of this specification. For ordinary technicians in this field, without paying creative work, the technical solutions or means disclosed in this specification can also be applied to other scenarios based on these technical contents.

[0019] It should be understood that the "system", "device", "unit" and / or "module" used in this specification is a method for distinguishing different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

[0020] Unless otherwise specified, technical terms used in this specification to describe components, elements, etc. do not refer to the singular, but may also include the plural. Generally speaking, terms such as "include", "comprise", etc. only indicate that the steps, elements, or components that have been clearly identified are included, and these steps, elements, and components do not constitute an exclusive list, such as the method or device described may also include other steps or components.

[0021] Flowcharts are used in this specification to illustrate the operation steps performed by the device or system of the relevant embodiments, but unless otherwise specified, the order used to describe these steps should not be understood as a limitation on the order in which the steps are performed. A person of ordinary skill in the art can adjust the order in which these steps are performed based on the knowledge and information conveyed by the embodiments of this specification, and the adjustment includes but is not limited to swapping the order, merging multiple steps, and splitting a certain step.

[0022] Virtual games, also known as video games or electronic games, are games that run on electronic devices and are controlled by computer programs (such as game applications) for educational or entertainment purposes. The game scenes of virtual games can be presented through a graphical user interface (GUI). Game scenes (or virtual scenes) include elements such as the environment, characters, buildings, machinery, and props in virtual games, which can restore at least part of the virtual world in the game. Users can input game operations or game control instructions through human-computer interface devices (such as game controllers, microphones, keyboards, mice, touch screens, cameras, etc.), and computer programs respond to game operations and present them to users through a graphical user interface.

[0023] Figure 1 Schematic diagram of a virtual game system according to some embodiments of this specification. Figure 1 As shown, the virtual game system 100 may include a server 110, a network 120, a storage medium 130, and a user terminal 140. The server 110 and the user terminal 140 may perform data transmission via the network 120.

[0024] The server 110 (or game server) may be a computer device with high computing performance, used to maintain and manage game data and game player data, synchronize and broadcast the interactive data of game players in the game, or complete some relatively complex calculations or logic processing in the virtual game. In some embodiments, the server 110 may also publish various versions of game applications for game players to download or update. In some embodiments, the server 110 may be a single computer device or a computing cluster composed of multiple computer devices, so as to provide more powerful computing power and more efficient response to service requests of game players.

[0025] The network 120 may be any form of wired or wireless network, or any combination thereof. As just an example, the network 120 may be one or more combinations of a wired network, an optical fiber network, a telecommunications network, an intranet, the Internet, a local area network (LAN), a wide area network (WAN), a wireless local area network (WLAN), a metropolitan area network (MAN), a wide area network (WAN), a public switched telephone network (PSTN), a Bluetooth network, etc. The network 120 may have multiple access points, and the server 110 and the user terminal 140 may access the network 120 through the access points.

[0026] The storage medium 130 may be used to store data and / or instructions related to the virtual game system 100. In some embodiments, the storage medium 130 may store data and / or information obtained from the server 110, the user terminal 140, etc. As an example, the storage medium 130 may store game player data, etc. In some embodiments, the storage medium 130 may include one or more storage components, each of which may be an independent device or part of other devices. In some embodiments, the storage medium 130 may be set in the server 110. In some embodiments, the storage medium 130 may include a random access memory (RAM), a read-only memory (ROM), a large capacity memory, a removable memory, a volatile read-write memory, etc., or any combination thereof. Exemplarily, the large capacity storage may include a magnetic disk, an optical disk, a solid-state disk, etc. In some embodiments, the storage medium 130 may be implemented on a cloud platform.

[0027] The user terminal (or terminal device) 140 may include but is not limited to a desktop computer, a smart phone, a laptop computer, a VR (Virtual Reality) device, a game console, and a tablet computer. Game players can install and run game applications on the user terminal 140, display game screens through the user terminal, and participate in the game through a human-computer interface device. As an example, the user terminal 140 may have a signal connection with the server 110 through the network 120, and the game player may download the game from the server 110 for installation. The human-computer interface device may be a part of the user terminal 140, or may be separated from the user terminal 140, but may have a signal connection with each other. In some embodiments, the virtual game may be a networked game, and the user terminals 140 of multiple game players may be connected to the server 110 through the network 120. With the participation of the server 110, multiple game players may interactively participate in the game.

[0028] It should be noted that Figure 1The scene diagram of the virtual game system shown is only an example. The virtual game system and scene described in the embodiment of this specification are to more clearly illustrate the technical solution of the embodiment of this specification, and do not constitute a limitation on the technical solution provided by the embodiment of this specification. For example, in a stand-alone game scene, Figure 1 The server 110, network 120 and storage medium 130 shown in the figure may be omitted. A person skilled in the art will appreciate that, with the evolution of game control systems and the emergence of new business scenarios, the technical solutions provided in the embodiments of this specification are also applicable to similar technical problems.

[0029] The embodiments of this specification provide a method, system and computer program product for prompting game information. Specifically, the method for prompting game information in the embodiments of this specification can be executed by a computer device, wherein the computer device can be a terminal (such as a user terminal) or a server. The terminal can be a terminal device such as a smart phone, a tablet computer, a laptop computer, a touch screen, a game console, a personal computer (PC, Personal Computer), a personal digital assistant (Personal Digital Assistant, PDA), etc. The terminal can also include a client, which can be a game application client, a browser client carrying a game program, or an instant messaging client, etc. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Distribute Network, content distribution network), and big data and artificial intelligence platforms.

[0030] In some embodiments, the virtual game may be any of shooting games, competitive games, casual games, action games, role-playing games, strategy games, sports games, puzzle games, and the like. Among them, the game may include a virtual scene of the game (or referred to as a game scene). The virtual scene of the game may include one or more game tasks, such as collection tasks, combat tasks, escape tasks, or main tasks that promote the progress of the game. The virtual scene of the game may also include one or more virtual characters (or referred to as game characters), and one or more virtual characters may be controlled by a user (or player) or by a computer, such as a robot of an artificial intelligence (AI) algorithm, to replace a real player to control a virtual character to play the game. The virtual scene of the game may also include game terrain, such as natural terrains such as mountains, plains, rivers, gullies, lakes, and oceans, or artificial terrains such as cities, villages, and buildings. The virtual scene of the game may also include virtual objects, such as railings, walls, doors and windows, plants, rocks, and the like. Some game terrains and virtual objects may be used as virtual shelters to provide shielding and protection functions for virtual characters, such as buildings, trees, rocks, walls, ridges, gullies, and the like. The virtual scene of the game may also include virtual items, such as food, medical items, ammunition, equipment, protective gear, firearms, tools or clues related to game tasks, etc. Some virtual items can also be collected and used by virtual characters as resources. The virtual scene of the game may also include one or more elements, such as skills, scores, character health status (such as life value, protective gear status, virtual character injury status, hunger value, thirst value, etc.), energy, etc., to provide help to players, provide virtual services, increase scores related to player performance, etc. In addition, the virtual scene of the game may also include one or more prompt information, which may be output in the form of text, voice, icon, color change, image / video, gesture, dialog box, vibration feedback, sound effect, etc. The content of some prompt information may be prompt information related to game tasks, path finding paths or combat conditions, to help players choose appropriate game tasks, path finding paths or understand combat conditions, etc. In some embodiments, the virtual scene of the game may also include a game scene that provides a game environment. A game may be an activity in which two or more virtual characters compete against each other in the game. Players controlling the virtual characters may achieve specific goals or win the game by developing strategies or techniques in the game. The content included in the virtual scene of the game is also applicable to the game scene.

[0031] In some embodiments, for more complex virtual games, such as those that require exploring maps to collect resources, or shooting games that include battle scenes, players (or users) need to have rich experience and skills to achieve good results in the game. However, for ordinary players who lack experience, especially novice players, how to plan game tasks, design game paths, or formulate effective combat strategies is often a huge challenge. These players may fail frequently and find it difficult to play in depth because they do not understand the game mechanism, map characteristics, or enemy behavior patterns, resulting in poor gaming experience and thus giving up the game. In addition, the main way for players to accumulate experience is to increase the number of games played. Although this method is effective, it requires a high time cost and also reduces the gaming experience.

[0032] To this end, some embodiments of the present specification propose a game information prompt method, which determines the current position of the target game task and the first virtual character, and determines the path finding path from the current position to the designated position indicated by the target game task through a path planning model. In the process of the first virtual character moving along the path finding path to the designated position, task guidance information is generated in a graphical user interface according to the real-time map data corresponding to the first game map, so as to reduce the task difficulty of the first virtual character to complete the target game task, lower the game threshold, and enable players to quickly understand the situation in the game game scene, so that players can enter into a deep play state such as conceiving battle strategies and coordinated operations according to the task guidance information, thereby enhancing the player's gaming experience.

[0033] Figure 2 This is an exemplary flow chart of a game information prompt method shown in some embodiments of this specification. Figure 2 The process 200 shown may be executed by a terminal device, for example, Figure 1 The user terminal 140 shown is executed. Or, Figure 2 The process 200 shown may also be executed by other processing devices, such as Figure 1 In some embodiments, the process 200 may be implemented by a game information prompting system 1000 deployed on a processing device or a terminal device. Figure 2 As shown, process 200 may include the following steps.

[0034] Step 210 , determining the target game task and the current position of the first virtual character. In some embodiments, step 210 may be implemented by the determination module 1010 .

[0035] In some embodiments, the terminal device may provide a graphical user interface, in which at least part of the game scene is displayed, and the game scene may include a first virtual character participating in the game and a second virtual character in a different game camp from the first virtual character. The first virtual character may be controlled by the terminal device, for example, the player may control the first virtual character to move, jump, crawl, pick up items, aim and shoot, etc. through the user terminal 140. The second virtual character may be controlled by a computer, such as a robot based on an artificial intelligence (AI) algorithm, to control the second virtual character to play the game instead of a real player, and the second virtual character may also be controlled by other players through other terminal devices.

[0036] In some embodiments, the target game task is used to instruct the first virtual character to complete a specified game behavior at a specified position in the game match scene. Among them, the specified position can be represented by the coordinate value in the world coordinate system corresponding to the first game map, and the current position of the first virtual character can also be represented by the coordinate value of the first virtual character in the world coordinate system. The world coordinate system can be a global coordinate system used to describe and define the position of virtual objects in the virtual scene in the virtual game. The first game map can correspond to the game match. Specifically, the first game map can be a game map in the game match scene, including virtual terrain, virtual objects, etc. The first virtual character and the second virtual character can move in the first game map and interact with the virtual objects in the first game map, such as picking up virtual objects. Game behavior can include various operations or behaviors performed by the first virtual character in the game match scene, such as collecting resources, teleporting to other scenes, placing virtual items, completing specific tasks, etc. Exemplarily, the target game task can be an evacuation task, which is used to instruct the first virtual character to reach a specified position in the first map within a preset time and be successfully teleported to a virtual scene outside the game match.

[0037] In some embodiments, the target game task can be determined in a virtual scene corresponding to a non-game match. Exemplarily, a player can select a main task in a virtual scene corresponding to a non-game match, and the user terminal 140 determines the main task as the target game task accordingly. Further, after determining the target game task, the game corresponding to the target game task can be entered, and the game scene corresponding to the target game task is displayed in the graphical user interface of the terminal device accordingly. In other embodiments, the target game task can also be determined in a virtual scene corresponding to the game match. Exemplarily, a player can select a map or theme in a virtual scene corresponding to the game match, and the user terminal 140 determines the game task corresponding to the map or theme as the target game task.

[0038] Step 220 , determine the path from the current location to the specified location through the path planning model, and display it in the graphical user interface. In some embodiments, step 220 can be implemented by the path planning module 1020 .

[0039] In some embodiments, the path planning model can be implemented based on a machine learning model, such as a deep neural network, or based on a reinforcement learning model, such as a deep Q network (DQN) or a policy gradient method. According to a certain embodiment, the path planning model may include a multi-layer neural network, specifically, an input layer, a hidden layer, and an output layer. Among them, the input layer can receive input data, such as the current position of the first virtual character, the specified position in the target game task, the map data corresponding to the first game map, etc., and the hidden layer may include multiple layers, and the layers are fully connected. Full connection means that any neuron in each layer is connected to any neuron in the next layer. The hidden layer may include, for example, a convolutional layer, a recurrent layer, a fully connected layer, etc., and the convolutional layer processes grid data, such as identifying obstacles in the map, confirming whether the path is passable, etc., and the recurrent layer processes sequence data, such as the node sequence in the path, to capture time or sequence dependencies, etc., and the fully connected layer integrates the features extracted by the convolutional layer and the recurrent layer to further capture high-level feature combinations and generate the final decision output. The output layer can output the path-finding path from the current position to the specified position according to the processing results of the hidden layer.

[0040] In some embodiments, the path planning model can be obtained by training the first historical game match data corresponding to the first game map. Among them, the first historical game match data may include relevant historical data generated by the historical first virtual characters controlled by different players playing games in the game match corresponding to the first game map. Specifically, the path planning model can be obtained by training the historical path data in the first historical game match data corresponding to the first game map and at least one of the following data: whether the task is completed data corresponding to the historical path data, the casualty data corresponding to the historical path data, the task completion time corresponding to the historical path data, and the revenue data corresponding to the historical path data. Among them, the historical path data may include multiple historical paths corresponding to the position of the first virtual character to the designated position indicated in the target game task. The task completion data may be data on whether the target game task is completed, for example, there are multiple historical path data from the current position of the first virtual character to the designated position indicated in the target game task, and each historical path data corresponds to data on whether a target game task is completed. The casualty data may include the life value or health value corresponding to the historical first virtual character, for example, there are multiple historical path data from the preset starting point corresponding to when the historical first virtual character enters the first game map to the designated location indicated in the target game task, and each historical path data corresponds to the life value or health value corresponding to when the historical first virtual character leaves the game match. The task completion time may be the total time corresponding to the time from entering the game match to completing the target game task.

[0041] In some embodiments, the first historical virtual character can collect resources with value attributes in historical game matches, and these resources can be converted into specific benefits according to their value attributes. For example, the resources with value attributes can be virtual items such as food, medical supplies, ammunition, equipment, protective gear, firearms, etc. with preset values. The income data corresponding to the historical path data may include the total income corresponding to the resources with value attributes collected along the way by the first historical virtual character in the process of traveling along the historical path. For example, a total of n resources are collected in the process of traveling along path A, and each resource corresponds to a different preset value (such as 1,000 coins). The income data corresponding to the historical path can be the total value of the n resources calculated according to their respective preset values.

[0042] In the path planning model training stage, the collected first historical game match data can be preprocessed, such as denoising the data, extracting useful information, etc., and then feature extraction can be performed on the preprocessed data. The extracted features can be, for example, the position of each node on the historical path, the task completion status of the target game task, the life value or health value corresponding to the historical first virtual character, the task completion time of the target game task, the income status of the target game task, etc. Further, the corresponding labels are determined according to the first historical game match data, such as path quality (such as whether the path is of high quality) or path priority, etc., or multi-dimensional labels are introduced, such as whether the task is completed. The degree of casualties (such as mild, moderate and severe), the speed of task completion (such as fast, general and slow), and the level of benefits (such as high benefits, medium benefits and low benefits), etc. Further, the extracted features and corresponding labels are used as training samples to train the path planning model, and the prediction results output by the path planning model are obtained. The difference between the prediction results output by the path planning model and the labels is evaluated through the loss function (such as mean square error, cross entropy, etc.), and the model parameters of the path planning model (such as the weight parameters in the model) are updated through the gradient descent algorithm to minimize the loss function, and the preset number of training times or expected loss convergence is achieved through iterative training. Furthermore, the validation set or cross-validation can be used to evaluate the performance of the model and check the performance of the model during the training process. After completing the training and debugging, the trained path planning model is deployed in the virtual game for online testing and monitoring, and further optimized and adjusted according to the actual results. During the game, data in the game can be collected in real time through an embedded system or plug-in, such as the current position of the first virtual character, the specified position in the target game task, and the real-time map data corresponding to the first game map, and then input into the path planning model. The path planning model outputs a path-finding path based on the input data.

[0043] For example, the collected current position of the first virtual character and the designated position in the target game task can be input into the path planning model. The path planning model determines a path finding path with high target game task completion rate, low casualty rate, short time to complete the target game task or high benefits as a high-quality path, and outputs the path finding path.

[0044] In some embodiments, indication information of the navigation path can be displayed in the graphical user interface of the terminal device, for example, the position corresponding to the navigation path is highlighted in the graphical user interface by a specific color or a specific brightness, and for example, an arrow icon indicating the direction of the navigation path is displayed in the graphical user interface.

[0045] Step 230, in the process of the first virtual character moving along the pathfinding path to the designated location, task guidance information is generated in the graphical user interface according to the real-time map data corresponding to the first game map, so as to reduce the task difficulty of the first virtual character to complete the target game task. In some embodiments, step 230 can be implemented by the information guidance module 1030.

[0046] In some embodiments, the map data corresponding to the first game map may include at least one of the following data: map resource distribution, the role position of the second virtual character, the first role data corresponding to the first virtual character, the second role data corresponding to the second virtual character, the location of the virtual bunker, and the range of the safe area. Among them, resources are, for example, resources with value attributes, which can be converted into specific benefits according to their value attributes. Resources with value attributes can be, for example, virtual items such as food, medical supplies, ammunition, equipment, protective gear, and firearms with preset values. The map resource distribution may include the location of resources in the first game map, the number of resources, the type of resources, etc. The virtual bunker may be a virtual object or game terrain that provides shielding and protection functions. The role position of the second virtual character may be represented by a coordinate value in the world coordinate system. The range of the safe area may be a dynamically changing safe area in the first game map. When the first virtual character is in the safe area, the health status of the first virtual character is not affected. When the first virtual character is outside the safe area, the health status of the first virtual character will be affected to a certain extent. For example, the life value in the health status decreases at a certain frequency. Among them, the dynamically changing safety area may be, for example, a safety area defined by a shrinking process. The shrinking process refers to a process in which the safety area is gradually reduced based on a game setting mechanism during the game. For example, the range defined by a circle determined by a dynamic radius with a set position in the first game map as the center is the safety range. The dynamic radius is periodically reduced according to the game setting mechanism, so that the range of the safety area is dynamically reduced. The character data includes at least one of health, equipment, and supply resources. Exemplarily, health may include life value (or health value), hunger value, thirst value, etc., equipment may include weapons (such as daggers, firearms, grenades, etc.), armor (such as helmets, chest armor, wrist guards, etc.), etc., and supply resources may include potions, food, ammunition, etc. The real-time map data corresponding to the first game map may be the map data corresponding to the first game map when the first virtual character is playing in a game match. The historical map data corresponding to the first game map may be the map data corresponding to the first game map when the historical first virtual character controlled by different players plays in a historical game match.

[0047] In some embodiments, the real-time map data corresponding to the first game map can be dynamically updated. For example, each time the first virtual character enters a game match, the map resource distribution in the real-time map corresponding to the first game map, the character position of the second virtual character, or the safety area range may be different.

[0048] In some embodiments, the task guidance information includes at least one of the following data: game behavior recommendation information, movement route prompt information, warning event prompt information, and combat success rate prediction information. Among them, the target game task can be associated with one or more sub-game tasks, and the sub-game task is used to instruct the first virtual character to complete a specified sub-game behavior in the game game scene, such as collecting specified resources, placing specified virtual items at the location specified by the sub-game task, defeating a specified number of enemies, etc. The game behavior recommendation information may include information for changing the execution priority of the sub-game task to indicate that the first virtual character should give priority to completing certain specified sub-game behaviors in the game game scene. The movement route prompt information may include prompt information for adjusting the direction of travel of the first virtual character to flexibly adjust the direction of travel of the first virtual character according to real-time map data. Among them, the direction of travel may indicate that the first virtual character moves in a specific direction at the current position, such as indicating that the first virtual character moves forward, backward, left, right, etc. The warning event prompt information may include relevant warning events when the first virtual character is in a combat state. The combat success rate prediction information may include the probability of winning when the first virtual character is in a combat state.

[0049] The following specifically describes how to generate task guidance information in a graphical user interface based on the real-time map data corresponding to the first game map from four aspects: game behavior recommendation information, movement route prompt information, warning event prompt information, and combat success rate prediction information.

[0050] Figure 3 This is an exemplary flow chart of a method for determining gaming behavior recommendation information according to some embodiments of this specification. Figure 3 The process 300 shown can be executed by a terminal device or a processing device. In some embodiments, the process 300 can be implemented by a task recommendation unit 1031 in the game information prompt system 1000 deployed on a processing device or a terminal device. Figure 3 As shown, process 300 may include the following steps.

[0051] Step 310, adjusting the execution priority of the sub-game task based on the task attributes of the sub-game task related to the target game task and at least one of the following data in the real-time map data corresponding to the first game map through the game task recommendation model: the current position of the first virtual character, the character position of the second virtual character, and the map resource distribution.

[0052] In some embodiments, the task attributes of the sub-game task may include a task type and a task location. The task type may be, for example, a resource collection type, a combat type, an escape type, etc. The task location may be represented by a coordinate value in a world coordinate system.

[0053] In some embodiments, the architecture of the game task recommendation model is similar to that of the path planning model, and specific reference may be made to the description in the above step 220. In other embodiments, the game task recommendation model may further include a calculation unit, which may further calculate the data input to the input layer and obtain a calculation result, and then input the calculation result to the hidden layer, which processes the data based on the data of the input layer and the calculation result of the calculation unit and outputs the processing result to the output layer, and the output layer outputs the final result.

[0054] According to one embodiment, the calculation unit may be used to calculate the execution distance corresponding to each sub-game task. For example, the calculation unit may calculate the distance between the task location of each sub-game task and the current position of the first virtual character based on the Euclidean distance function.

[0055] According to another embodiment, the calculation unit can also be used to calculate the security corresponding to each sub-game task. Specifically, the calculation unit can determine the security of each sub-game task based on the number of second virtual characters within the corresponding range of the task location of each sub-game task, wherein the corresponding range of the task location of the sub-game task can be a first preset range determined with the task location as the center, and the security of the sub-game task can be set by setting multiple quantity thresholds. Exemplarily, the more the number of second virtual characters falling into the first preset range, the lower the security of the sub-game task.

[0056] According to another embodiment, the calculation unit can also be used to calculate the estimated income corresponding to the sub-game tasks. Specifically, the calculation unit can determine the estimated income of each sub-game task based on the number of resources within the corresponding range of the task location of the sub-game task and the preset value corresponding to the resources, wherein the corresponding range of the task location of the sub-game task can be a second preset range determined with the task location as the center, and the second preset range can be the same as the first preset range or different, and further, the total value of the resources within the second preset range can be calculated based on the preset value (such as 1000 coins) corresponding to each resource within the second preset range, and used as the estimated income of the sub-game task.

[0057] In some embodiments, the game task recommendation model can be trained by the second historical game match data corresponding to the first game map. Among them, the second historical game match data may include the task attributes of the historical sub-game task and at least one of the following data: the behavior data corresponding to the historical first virtual character, the task completion data corresponding to the historical sub-game task, the task completion time corresponding to the historical sub-game task, and the historical map data corresponding to the first game map. Among them, the historical first virtual character may be the historical first virtual character controlled by different players. The behavior data corresponding to the historical first virtual character may include the task execution data of the historical first virtual character in the historical game match corresponding to the first game map (such as the operation of switching sub-game tasks, the order of executing sub-game tasks, etc.). The historical map data corresponding to the first game map may include the position of the historical first virtual character, the role position of the historical second virtual character, and the historical map resource distribution corresponding to the first game map.

[0058] In the game task recommendation model training stage, the collected second historical game match data can be preprocessed, such as denoising the data, extracting useful information, etc., and then feature extraction can be performed on the preprocessed data. The extracted features can be, for example, task attributes of historical sub-game tasks, the execution order of the historical first virtual character on the historical sub-game tasks, the task completion status of each historical sub-game task, the task completion time of each historical sub-game task, the resource collection status corresponding to each historical sub-game task, the historical resource distribution in the first game map, the position of the historical first virtual character, the role position of the historical second virtual character, etc. Corresponding labels are determined according to the second historical game match data, such as the execution priority of the historical sub-game tasks. Further, the extracted features and corresponding labels are used as training samples. The game task recommendation model is trained, and the input features are transmitted to the calculation unit through the input layer of the game task recommendation model. The calculation unit calculates the execution distance corresponding to the historical sub-game task, the security corresponding to the historical sub-game task, the estimated income corresponding to the historical sub-game task, etc. according to the input features, and outputs the calculation results to the hidden layer. The hidden layer processes according to the calculation results of the calculation unit and the input features of the input layer, and outputs the prediction results through the output layer. The difference between the prediction results and the labels output by the game task recommendation model is further evaluated through the loss function (such as mean square error, cross entropy, etc.), and the model parameters of the game task recommendation model (such as the weight parameters in the model) are updated through the gradient descent algorithm to minimize the loss function. The preset number of training times or the expected loss convergence is achieved through iterative training. Furthermore, the validation set or cross-validation can be used to evaluate the performance of the model and check the performance of the model during the training process. After completing the training and debugging, the trained game task recommendation model is deployed in the virtual game for online testing and monitoring, and further optimized and adjusted according to the actual effect.

[0059] During the game, in-game data can be collected in real time through an embedded system or plug-in, such as the task locations of each sub-game task related to the target game task, the current position of the first virtual character, the role position of the second virtual character, and the distribution of map resources in real-time map data, and input into the trained game task recommendation model. The game task recommendation model outputs the execution priority of the sub-game task based on the input data.

[0060] In some embodiments, the game task recommendation model can determine the execution distance corresponding to the sub-game task based on the task location and the current location of the first virtual character, and adjust and output the execution priority of the sub-game task based on the execution distance. Exemplarily, the shorter the execution distance of the sub-game task, the higher the execution priority of the sub-game task.

[0061] In some embodiments, the game task recommendation model can also determine the security of the sub-game task based on the role position of the second virtual character and the task location of the sub-game task, and adjust the execution priority of the sub-game task based on the security. Exemplarily, the higher the security of the sub-game task, the higher the execution priority of the sub-game task.

[0062] In some embodiments, the game task recommendation model can also determine the estimated revenue corresponding to the sub-game task based on the task location and map resource distribution of the sub-game task, and adjust the execution priority of the sub-game task based on the estimated revenue. For example, the higher the estimated revenue corresponding to the sub-game task, the higher the execution priority of the sub-game task.

[0063] Step 320: Displaying game behavior recommendation information for changing the execution priority of the sub-game task in the graphical user interface according to the execution priority of the sub-game task.

[0064] In some embodiments, the sub-game tasks to be recommended may be determined based on the execution priority of the sub-game tasks, and corresponding game behavior recommendation information may be generated based on the sub-game tasks to be recommended. The game behavior recommendation information may include text information, which is suitable for prompting the player whether to execute the sub-game tasks to be recommended, or prompting the player whether to change the sub-game tasks being executed to the sub-game tasks to be recommended. Figure 4 is a schematic diagram of a graphical user interface according to some embodiments of this specification. For example, Figure 4As shown in the figure, the sub-game task to be recommended is "obtain the key to safe No. 5", then the game behavior recommendation information "Current status is good, equipment is in the top 20% level among competitors in the same game, it is recommended to prioritize the task of 'obtain safe No. 5'" can be displayed in the graphical user interface. Furthermore, the preset display time of the game behavior recommendation information can be set, and the displayed time of the game behavior recommendation information can be presented in the form of a progress bar. For example, Figure 4 As shown, the preset display time of the game behavior recommendation information is indicated by the length of the progress bar below the aforementioned game behavior recommendation information, and the displayed time of the game behavior recommendation information is indicated by the change of the progress bar color. The game behavior recommendation information "Are you sure to change the task priority and start smart navigation?" can be further displayed around the progress bar. The player can confirm the execution of the sub-game task to be recommended by performing corresponding operations in the graphical user interface or voice reply. For example, when the player clicks the task "Get the key to safe No. 5" in the graphical user interface or voice reply "Confirm", the execution of the sub-game task to be recommended is confirmed.

[0065] Figure 5 This is an exemplary flowchart of a method for determining mobile route prompt information according to some embodiments of this specification. Figure 5 The process 500 shown can be executed by a terminal device or a processing device. In some embodiments, the process 500 can be implemented by the direction adjustment unit 1032 in the game information prompt system 1000 deployed on the processing device or the terminal device. Figure 5 As shown, process 500 may include the following steps.

[0066] Step 510, determining the direction of travel of the first virtual character through a direction adjustment model based on at least one of the following data in the real-time map data corresponding to the first game map: the current position of the first virtual character, the character position of the second virtual character, the second character data corresponding to the second virtual character, the map resource distribution, and the range of the safe area.

[0067] In some embodiments, the architecture of the direction adjustment model may be similar to the architecture of the game task recommendation model. For details, please refer to the description in the above step 310. The difference lies in that the content calculated by the calculation unit is different.

[0068] According to a certain embodiment, the calculation unit of the direction adjustment model can be used to calculate the collectible resources corresponding to the path from the current position of the first virtual character to the task location corresponding to the sub-game task being executed. Specifically, multiple alternative paths from the current position of the first virtual character to the task location corresponding to the sub-game task being executed can be determined based on the A* algorithm, Dijkstra algorithm or other path planning algorithms, and the range corresponding to each alternative path can be determined based on the waypoints on the alternative path and the set distance. In some embodiments, the range corresponding to the alternative path can be a set of areas enclosed by a preset geometric shape with one or more waypoints as the geometric center and a set distance as the radius or side length. Exemplarily, the range corresponding to the alternative path is a set of circular areas with one or more waypoints as the center and a set distance as the radius. Further, the collectible resources within the corresponding range of each alternative path are determined based on the map data. Among them, the collectible resources can be resources that can be collected or used by the first virtual character in the first game map. Exemplarily, the position of the collectible resources within the corresponding range of each alternative path can be determined based on the resource distribution in the map data.

[0069] According to a certain embodiment, the calculation unit of the direction adjustment model can also be used to determine the danger level information within the range corresponding to the current position of the first virtual character based on the position of the second virtual character and the second character data corresponding to the second virtual character. Specifically, the range corresponding to the current position of the first virtual character can be a third preset range determined with the current position of the first virtual character as the center, and the danger level corresponding to the current position of the first virtual character is determined based on the number of second virtual characters within the third preset range and the second character data. For example, the more the number of second virtual characters falling within the first preset range, or the higher the health value or the higher the equipment level in the second character data corresponding to the second virtual character, the higher the danger level corresponding to the current position of the first virtual character.

[0070] In some embodiments, the direction adjustment model can be trained by the third historical game match data corresponding to the first game map, and the third historical game match data includes the behavior data corresponding to the historical first virtual character and at least one of the following data: resource collection data of the historical first virtual character, and historical map data corresponding to the first game map. Among them, the character data includes at least one of health, equipment, and supply resources. The behavior data corresponding to the historical first virtual character may include the behavior trajectory data of the historical first virtual character in the historical game match corresponding to the first game map (such as movement path, movement direction, residence time in a specific location or area, etc.), action interaction data (such as using weapons or equipment, using skills, picking up items, trading items, etc.), state and attribute change data (such as health or hunger value change data, supply resource change data in the backpack), etc. The resource collection data of the historical first virtual character may include the resources collected or used by the historical first virtual character during the historical game match. The historical map data corresponding to the first game map may include the position of the historical first virtual character, the role position of the historical second virtual character, the second role data corresponding to the historical second virtual character, the historical map resource distribution corresponding to the first game map, and the historical safety area range corresponding to the first game map.

[0071] In the direction adjustment model training stage, the collected third historical game match data can be preprocessed, such as denoising the data, extracting useful information, etc., and then feature extraction can be performed on the preprocessed data. The extracted features can be, for example, the behavior trajectory of the first historical virtual character, the resource collection status of the first historical virtual character, the position of the first historical virtual character, the role position of the second historical virtual character, the second role data corresponding to the second historical virtual character, the resource distribution of the historical map corresponding to the first game map, the historical safety area range corresponding to the first game map, etc., and the corresponding labels are determined according to the third historical game match data, such as whether the direction of travel of the first historical virtual character is accurate. Further, the extracted features and corresponding labels are used as training samples to train the direction adjustment model. The input layer of the direction adjustment model outputs the input features to the calculation unit and the hidden layer. The calculation unit calculates the direction adjustment data from the first virtual character according to the input features. The calculation results are output to the hidden layer, which processes the calculation results of the calculation unit and the input features of the input layer, and outputs the prediction results (such as the predicted direction of travel) through the output layer. The difference between the prediction results output by the direction adjustment model and the label is further evaluated through the loss function (such as mean square error, cross entropy, etc.), and the model parameters of the direction adjustment model (such as weight parameters in the model) are updated through the gradient descent algorithm to minimize the loss function, and the preset number of training times or expected loss convergence is achieved through iterative training. Furthermore, the validation set or cross-validation can be used to evaluate the performance of the model and check the performance of the model during the training process. After completing the training and debugging, the trained direction adjustment model is deployed in the virtual game for online testing and monitoring, and further optimized and adjusted according to the actual results.

[0072] During the game, the embedded system or plug-in can collect data in the game in real time, such as the task location corresponding to the sub-game task, the current position of the first virtual character, the role position of the second virtual character, the second role data corresponding to the second virtual character, the map resource distribution, the safe area range, etc., and input them into the trained direction adjustment model. The direction adjustment model adjusts and outputs the direction of travel of the first virtual character based on the input data. In the process of moving to the specified position along the path finding path, the current position of the first virtual character is constantly changing, and the role position and safe area range of the second virtual character can also be updated in real time. The direction adjustment model needs to be continuously updated and output the calculation results according to the real-time changing data, so as to provide effective direction guidance information to the player in real time.

[0073] In some embodiments, in response to an instruction to change a sub-game task, a direction of travel from the current position of the first virtual character to a task location corresponding to the changed sub-game task may be determined by a direction adjustment model, wherein the sub-game task is related to the target game task. Specifically, the sub-game task may be changed based on a player input operation. For example, when the player is in a state such as Figure 4 In the graphical user interface shown, click on the task "Get the key to safe No. 5" to confirm the change of the sub-game task being executed. Further, when the sub-game task is changed, the task location corresponding to the sub-game task input into the direction adjustment model changes accordingly, and the direction adjustment model recalculates and outputs the recalculated moving direction of the first virtual character according to the task location corresponding to the updated sub-game task and the current position of the first virtual character.

[0074] In some embodiments, the direction adjustment model can be used to determine the collectible resources corresponding to the path from the current position of the first virtual character to the task location corresponding to the sub-game task being executed based on the map resource distribution, and the direction of travel of the first virtual character can be adjusted according to the resource location of the collectible resources. Exemplarily, there can be multiple alternative paths from the current position of the first virtual character to the task location corresponding to the sub-game task being executed, and the direction adjustment model can adjust the direction of travel of the first virtual character based on the collectible resources within the corresponding range of each alternative path, so that the direction of travel of the first virtual character points to the alternative path with the largest number of collectible resources within the corresponding range of the alternative paths.

[0075] In some embodiments, the direction adjustment model can be used to determine the danger level information within the range corresponding to the current position of the first virtual character based on the position of the second virtual character and the second character data corresponding to the second virtual character, and the direction of travel of the first virtual character can be adjusted according to the danger level information. Exemplarily, the more the number of second virtual characters within the range corresponding to the current position of the first virtual character, or the higher the health value or the higher the equipment level of the second virtual characters within the range corresponding to the current position of the first virtual character, the higher the danger level within the range corresponding to the current position of the first virtual character, and the direction adjustment model can adjust the direction of travel of the first virtual character based on the danger level corresponding to the current position of the first virtual character, so that the direction of travel of the first virtual character is away from the position of the second virtual character within the range corresponding to the current position of the first virtual character.

[0076] In some embodiments, the direction of travel of the first virtual character pointing to the safe area range can be determined by the direction adjustment model based on the safe area range and the current position of the first virtual character. Exemplarily, the safe area range can be dynamically updated, and the direction of travel of the first virtual character can be adjusted by the direction adjustment model so that the direction of travel of the first virtual character points to the center of the safe area range.

[0077] Step 520: Displaying moving route prompt information for adjusting the moving direction of the first virtual character in the graphical user interface according to the moving direction of the first virtual character.

[0078] In some embodiments, the moving route prompt information for adjusting the moving direction of the first virtual character can be determined based on the moving direction of the first virtual character output by the direction adjustment model. The moving route prompt information can be identification information corresponding to the moving direction of the first virtual character, and the identification information corresponding to the moving direction can include text, icons, images or color information. Exemplarily, specific color information can be drawn at a path position corresponding to the moving direction of the first virtual character in a graphical user interface to identify the moving direction of the first virtual character. Figure 6 FIG. 1 is another schematic diagram of a graphical user interface according to some embodiments of this specification. Figure 6 As shown, an arrow icon corresponding to the direction of travel of the first virtual character may also be displayed in the graphical user interface to indicate the direction of travel of the first virtual character. Furthermore, corresponding text information "10 meters ahead, turn left" may also be displayed above the arrow icon.

[0079] In some embodiments, the moving route prompt information further includes at least one of the following information: danger information related to the path finding path, estimated benefit information related to the path finding path, estimated time related to the path finding path, and alternative bunkers related to the path finding path. The following describes how to generate the moving route prompt information related to the path finding path in the graphical user interface according to the real-time map data corresponding to the first game map from the above aspects.

[0080] In some embodiments, the danger level information related to the path-finding path can be displayed in a graphical user interface based on the character position of the second virtual character. Specifically, the range corresponding to one or more waypoints can be determined based on the waypoints on the path-finding path and the set distance, and the danger level can be determined based on the number of second virtual characters within the range corresponding to the one or more waypoints, thereby determining the danger level information related to the path-finding path. Exemplarily, the more the number of second virtual characters, the higher the danger level. Furthermore, the path-finding path can be segmented, each path including one or more waypoints, and the danger level corresponding to the path segment can be determined based on the number of second virtual characters within the range corresponding to one or more waypoints in each path segment, and the identification information of the path-finding path can be determined based on the danger level corresponding to each path segment. Exemplarily, different colors can be used to identify the danger level corresponding to each path segment.

[0081] In some embodiments, estimated revenue information related to the pathfinding path may also be displayed in the graphical user interface based on the map resource distribution. Specifically, the range corresponding to the pathfinding path may be determined based on the waypoints on the pathfinding path and the set distance, and the total estimated revenue of the resources within the range corresponding to the pathfinding path may be calculated based on the resource distribution within the range corresponding to the pathfinding path and the preset value corresponding to one or more resources, thereby determining the estimated revenue information related to the pathfinding path.

[0082] In some embodiments, the estimated time to reach the designated location can also be displayed in the graphical user interface based on the current location of the first virtual character, the designated location, and the moving speed of the first virtual character. Specifically, the task distance can be determined based on the current location of the first virtual character and the designated location corresponding to the target game task, and the estimated time to reach the designated location corresponding to the target game task can be determined based on the ratio of the task distance to the moving speed of the first virtual character.

[0083] In some embodiments, alternative shelter information related to the path finding path can also be displayed in the graphical user interface based on the position of the virtual shelter and the alternative shelter can be highlighted. Exemplarily, for each of the one or more waypoints on the path finding path, the shelter closest to the waypoint is used as an alternative shelter.

[0084] For example, Figure 6 As shown, the moving route prompt information related to the path-finding path generated in the graphical user interface based on the real-time map data corresponding to the first game map includes "danger level: low", "estimated income: 5000000", and "estimated time: 10 minutes".

[0085] Figure 7 This is an exemplary flowchart of another method for determining moving route prompt information according to some embodiments of this specification. Figure 7 The process 700 shown can be executed by a terminal device or a processing device. In some embodiments, the process 700 can be implemented by a game information prompt system 1000 deployed on a processing device or a terminal device. Figure 7 As shown, process 700 may include the following steps.

[0086] Step 710 , determining the required resources of the first virtual character based on the map resource distribution and the first character data by using the resource management model. In some embodiments, step 710 may be implemented by the resource management unit 1033 .

[0087] In some embodiments, the resource management model can be implemented based on a machine learning model, such as a deep neural network, or based on a reinforcement learning model, such as a deep Q network (DQN) or a policy gradient method. The architecture of the resource management model is similar to that of the path planning model. For details, please refer to the description in step 220 above, which will not be repeated here.

[0088] In some embodiments, the resource management model can be trained by the fourth historical game match data corresponding to the first game map. Among them, the fourth historical game match data includes the resource collection data of the historical first virtual character and the historical map data corresponding to the first game map. The resource collection data of the historical first virtual character may include the resources collected or used by the historical first virtual character during the game, and the historical map data corresponding to the first game map may include the historical map resource distribution corresponding to the first game map, and the first character data (such as supply resources) of the historical first virtual character. Among them, the supply resources may be the resources in the backpack of the first virtual character, which can be collected or used by the first virtual character, such as food resources, combat-related resources, and resources that enhance the life value or ability of the virtual character. Exemplarily, the supply resources may include potions that can enhance the life value or ability (such as attack power) of the first virtual character, food that can reduce the hunger value of the first virtual character, and ammunition that can enhance the combat effectiveness of the first virtual character. The demand resources may be supply resources that the first virtual character holds less or supply resources that are urgently needed. Accordingly, the demand resources may include at least one of the following data: food resources, combat-related resources, and resources that enhance the life value or ability of the virtual character.

[0089] In the resource management model training stage, the collected fourth historical game match data can be preprocessed, such as denoising the data, extracting useful information, etc., and then feature extraction can be performed on the preprocessed data. The extracted features can be, for example, the resource collection situation of the first virtual character in history, the supply resource situation of the first virtual character in history, the distribution of historical resources in the first game map, etc., and the corresponding labels can be determined according to the fourth historical game match data, such as whether the required resources of the first virtual character in history are accurate. The extracted features and corresponding labels are used as training samples to train the resource management model, and the prediction results output by the resource management model (such as the predicted required resources) are obtained. Further, the difference between the prediction results output by the resource management model and the labels is evaluated by a loss function (such as mean square error, cross entropy, etc.), and the model parameters of the resource management model (such as the weight parameters in the model) are updated by a gradient descent algorithm to minimize the loss function, and the preset number of training times or the expected loss convergence is achieved through iterative training. Further, a validation set or cross-validation can be used to evaluate the performance of the model and check the performance of the model during the training process. After training and debugging, the trained resource management model will be deployed to the virtual game for online testing and monitoring, and further optimization and adjustment will be made based on the actual results.

[0090] During the game, the embedded system or plug-in can collect data in the game in real time, such as the map resource distribution in the real-time map data and the supply resource data in the first character data, and input them into the trained resource management model, and the resource management model determines the required resources of the first virtual character based on the input data. Exemplarily, when the backpack of the first virtual character lacks a certain type of supply resource or the storage capacity of such supply resource is less than the first preset storage capacity, and the resource distribution information of the real-time map data indicates the existence of such supply resource, the resource management model determines such supply resource as the required resource of the first virtual character.

[0091] Step 720 , generating moving route prompt information for adjusting the moving direction of the first virtual character in the graphical user interface based on the required resources of the first virtual character and the map resource distribution through the direction adjustment model. In some embodiments, step 720 may be implemented by the direction adjustment unit 1032 .

[0092] In some embodiments, the training data of the direction adjustment model may further include the demand resource status of the first virtual character in history. The direction adjustment model may also be trained based on the demand resource status of the first virtual character in history. The training method may refer to the description in the previous step 510 and will not be repeated here.

[0093] During the game, the required resources of the first virtual character output by the resource management model can be input into the direction adjustment model, and the direction adjustment model adjusts and outputs the direction of travel of the first virtual character based on the required resources of the first virtual character and the map resource distribution in the real-time map data corresponding to the first game map collected by the embedded system or the plug-in. Exemplarily, assuming that the required resources of the first virtual character are combat-related resource A, the direction adjustment model can adjust and output the direction of travel of the first virtual character based on resource A and the location of resource A in the real-time map data, so that the direction of travel of the first virtual character points to the location of resource A in the real-time map data.

[0094] Figure 8 It is an exemplary flow chart of a method for determining warning event prompt information according to some embodiments of this specification. Figure 8 The process 800 shown can be executed by a terminal device or a processing device. In some embodiments, the process 800 can be implemented by the battle prediction unit 1034 in the game information prompt system 1000 deployed on the processing device or the terminal device. Figure 8 As shown, process 800 may include the following steps.

[0095] Step 810, determining whether the first virtual character enters a combat state according to preset conditions.

[0096] In some embodiments, the combat distance can be determined based on the current position of the first virtual character and the position of the second virtual character, and when the combat distance is less than a preset distance, the first virtual character is judged to have entered a combat state. Exemplarily, the current position of the first virtual character and the position of one or more second virtual characters can be obtained from the real-time map data corresponding to the first game map, and the combat distance between the current position of the first virtual character and the position of the nearest second virtual character is calculated, and when the combat distance is less than the preset distance, the first virtual character is judged to have entered a combat state.

[0097] In some embodiments, when the health value in the first character data decreases, it is determined that the first virtual character enters a combat state. Exemplarily, the health value in the health state of the first virtual character can be periodically obtained, and when the health value decreases, it is determined that the first virtual character enters a combat state.

[0098] In some embodiments, when the first virtual character performs a preset operation, it is determined that the first virtual character enters a combat state. Exemplarily, when the first virtual character performs an aiming operation or a shooting operation, it is determined that the first virtual character enters a combat state.

[0099] Step 820, in response to the first virtual character entering a combat state, generating warning event prompt information in a graphical user interface based on real-time map data corresponding to the first game map through a combat prediction model.

[0100] In some embodiments, the battle prediction model can be implemented based on a machine learning model, such as a deep neural network, or based on a reinforcement learning model, such as a deep Q network (DQN) or a policy gradient method. The architecture of the battle prediction model is similar to that of the path planning model. For details, please refer to the description in step 220 above, which will not be repeated here.

[0101] In some embodiments, the battle prediction model can be trained by the fifth historical game match data corresponding to the first game map. Among them, the fifth historical game match data includes the behavior data corresponding to the first historical virtual character and at least one of the following data: the resource collection data of the first historical virtual character, the historical map data corresponding to the first game map. Among them, the behavior data corresponding to the first historical virtual character may include the behavior trajectory data of the first historical virtual character in the historical game match corresponding to the first game map (such as the moving path, moving direction, the stay time in a specific location or area, etc.), action interaction data (such as using weapons or equipment, using skills, picking up items, trading items, etc.), state and attribute change data (such as the change data of health value or hunger value, the change data of supply resources in the backpack), etc. The resource collection data of the first historical virtual character may include the resources collected or used by the first historical virtual character during the historical game match. The historical map data corresponding to the first game map may include the position of the first historical virtual character, the role position of the second historical virtual character, the first role data corresponding to the first historical virtual character, the second role data corresponding to the second historical virtual character, the historical map resource distribution corresponding to the first game map, the historical safety area range corresponding to the first game map, the historical virtual bunker position corresponding to the first game map, etc.

[0102] In the battle prediction model training stage, the collected fifth historical game match data can be preprocessed, such as denoising the data, extracting useful information, etc., and then feature extraction can be performed on the preprocessed data. The extracted features can be, for example, the position of the first historical virtual character, the position of the second historical virtual character, the position of the historical virtual bunker, the behavior trajectory of the first historical virtual character, the comparison information of the first character data of the first historical virtual character and the second character data of the second historical virtual character, etc., and different corresponding labels can be set. After determining the extracted features and the corresponding labels, the extracted features and the corresponding labels can be further used as training samples to train the battle prediction model. The battle prediction outputs the prediction results according to the input features, and the difference between the prediction results output by the battle prediction model and the labels is evaluated through the loss function (such as mean square error, cross entropy, etc.), and the model parameters of the battle prediction model (such as the weight parameters in the model) are updated through the gradient descent algorithm to minimize the loss function, and the preset number of training times or the expected loss convergence is achieved through iterative training. Furthermore, the performance of the model can also be evaluated by a validation set or cross validation to check the performance of the model during the training process. After training and debugging, the trained combat prediction model is deployed to the virtual game for online testing and monitoring, and further optimized and adjusted according to the actual results. During the game, the embedded system or plug-in can collect data in the game in real time, such as the current position of the first virtual character, the weapon used by the first virtual character, the first character data corresponding to the first virtual character, the character position of the second virtual character, the second character data corresponding to the second virtual character, the position of the virtual bunker in the real-time map data, the range of the safe area in the real-time map data, the map resource distribution in the real-time map data, etc. The combat prediction model outputs corresponding warning event prompt information based on the input data.

[0103] In some embodiments, the warning event prompt information includes at least one of the following information: combat position, target bunker, number of bunkers, terrain conditions, intensity of battle, and status of the second virtual character.

[0104] In some embodiments, the combat prediction model can be used to display warning event prompt information indicating the combat position and / or target bunker in a graphical user interface based on the current position of the first virtual character, the role position of the second virtual character, and the position of the virtual bunker. Specifically, the features extracted by the combat prediction model in the training phase can specifically include the position of the first virtual character in history, the role position of the second virtual character in history, the position of the virtual bunker in history, the behavior trajectory of the first virtual character in history, etc. The corresponding label can be whether the combat position and / or the target bunker are accurate. The extracted features and corresponding labels can be used as training samples and the combat prediction model can be trained according to the above training method, and the corresponding data is collected in real time during the game and input into the combat prediction model. The combat prediction model outputs a suitable combat position and / or target bunker according to the input data. Exemplarily, during the game, the current position of the first virtual character, the role position of the second virtual character, and the position of the virtual bunker are input into the combat prediction model, and the combat prediction model outputs the combat position and / or the target bunker position that is located between the current position of the first virtual character and the role position of the second virtual character, and is closest to the first virtual character. Fig. 9 is another graphical user interface schematic diagram according to some embodiments of this specification. Fig. 9 As shown, the location of the target shelter can be displayed in the graphical user interface and the target shelter and combat position can be highlighted, and Fig. 9 The corresponding prompt information "Available cover has been highlighted, and the recommended combat area has been highlighted" is generated in the display interface shown.

[0105] In some embodiments, the number of bunkers can also be displayed in a graphical user interface based on the position of virtual bunkers through a battle prediction model. Specifically, the features extracted by the battle prediction model during the training phase may include the historical position of the first virtual character and the historical position of the virtual bunker, and the corresponding labels may be the number of bunkers (such as high number of bunkers, medium number of bunkers, low number of bunkers, etc.). The extracted features and corresponding labels can be used as training samples and the battle prediction model can be trained according to the above training method. During the game, corresponding data can be collected in real time and input into the battle prediction model. The battle prediction model outputs the number of bunkers or the specific number of bunkers based on the input data. Exemplarily, the position of the virtual bunker in the real-time map data and the current position of the first virtual character can be input into the battle prediction model, and the battle prediction model outputs the number of bunkers. Fig. 9 As shown, "Cover Amount: Medium" may be displayed in the graphical user interface.

[0106] In some embodiments, the terrain conditions can be displayed in a graphical user interface based on the current position of the first virtual character and the role position of the second virtual character by a combat prediction model. Among them, the role position of the second virtual character can include the role position of the second virtual character within the first range, and the first range can be determined based on the current position of the first virtual character and the range of the weapon used by the first virtual character. Exemplarily, the first range can be the range enclosed by a circle with the current position of the first virtual character as the center and the range of the weapon used by the first virtual character as the radius. Specifically, the features extracted by the combat prediction model in the training stage can specifically include the position of the first virtual character in history, the weapon used by the first virtual character in history and the weapon range, and the role position of the second virtual character in history. The corresponding label can be the terrain conditions (such as height difference, or the evaluation result of the terrain conditions is high or low). The extracted features and corresponding labels can be used as training samples and the combat prediction model can be trained according to the above training method, and the corresponding data can be collected in real time during the game and input into the combat prediction model, and the combat prediction model outputs the terrain conditions according to the input data. For example, the current position of the first virtual character, the weapon used by the first virtual character and the corresponding weapon range, and the character position of the second virtual character can be input into the battle prediction model, and the battle prediction model outputs the evaluation result of the terrain condition or the corresponding height difference based on the current position of the first virtual character and the character position of the second virtual character within the range of the weapon used by the first virtual character. Fig. 9 As shown, “height difference: 3 m” and the evaluation result of the terrain condition “dominant evaluation: high” can be displayed in the graphical user interface.

[0107] In some embodiments, the intensity of the battle can be displayed in a graphical user interface through a battle prediction model based on at least one of the following information: the character position of the second virtual character, the comparison information of the first character data and the second character data, wherein the character data includes at least one of health, equipment, and supply resources. Among them, the character position of the second virtual character may include the character position of the second virtual character within a first range, and the first range may be determined based on the current position of the first virtual character and the range of the weapon used by the first virtual character. The comparison information of the first character data and the second character data may include whether the health value in the first character data is greater than the health value in the second character data, whether the equipment level in the first character data is higher than the equipment level in the second character data, whether the supply resources in the first character data are more than the supply resources in the second character data, etc. Specifically, the features extracted by the battle prediction model during the training phase may include the position of the first virtual character in history, the weapon used by the first virtual character in history and the weapon range, the character position of the second virtual character in history, and the comparison information between the first character data of the first virtual character in history and the second character data of the second virtual character in history. The corresponding label may be the intensity of the battle (for example, the intensity of the battle is high, medium, low, etc.). The extracted features and corresponding labels may be used as training samples and the battle prediction model may be trained according to the above training method. The corresponding data may be collected in real time during the game and input into the battle prediction model. The battle prediction model outputs the intensity of the battle based on the input data. Exemplarily, the current position of the first virtual character, the weapon used by the first virtual character, the character position of the second virtual character, and the comparison information between the first character data and the second character data may be input into the battle prediction model. The battle prediction model outputs the intensity of the battle in the current battle state based on the current position of the first virtual character and the character position of the second virtual character within the range of the weapon used by the first virtual character, and the comparison information between the first character data and the second character data. Figure 6 As shown, "Battle Intensity: Medium" can be displayed in the graphical user interface.

[0108] In some embodiments, the combat prediction model can be used to highlight the status of the second virtual character in the graphical user interface based on the character position and the second character data of the second virtual character. The status of the second virtual character includes at least one of the character position of the second virtual character, the area where the second virtual character is located, the equipment status of the second virtual character, and the health status of the second virtual character. Specifically, the features extracted by the combat prediction model in the training phase may include the historical position of the first virtual character, the weapon used by the first virtual character and the weapon range, the character position of the second virtual character in history, and the second character data of the second virtual character in history. The corresponding label may be whether the description of the status of the second virtual character is accurate, wherein the status of the second virtual character may include at least one of the character position of the second virtual character, the area where the second virtual character is located, the equipment status of the second virtual character, and the health status of the second virtual character. The extracted features and corresponding labels can be used as training samples and the combat prediction model can be trained according to the above training method. The corresponding data can be collected in real time during the game and input into the combat prediction model. The combat prediction model outputs the status of the second virtual character based on the input data. Exemplarily, the current position of the first virtual character, the weapon used by the first virtual character and the weapon range, the character position of the second virtual character and the second character data can be input into the combat prediction model. The combat prediction model outputs the status of the second virtual character. Exemplarily, Fig. 9 As shown, the second virtual character status may include "one enemy is detected under the stairs, the enemy has a level 1 helmet and level 1 armor."

[0109] In some embodiments, combat success rate prediction information can also be generated by the combat prediction model based on at least one of the following information: combat distance, bunker coverage, comparison information of the first character data and the second character data, and the role position of the second virtual character. Specifically, the features extracted by the combat prediction model in the training phase may include the combat distance between the historical first virtual character and the historical second virtual character, the historical bunker coverage, the comparison information of the first character data of the historical first virtual character and the second character data of the historical second virtual character, the role position of the historical second virtual character, the position of the historical first virtual character, the weapon used by the historical first virtual character, and the weapon range, and the corresponding label may be the combat success rate. Among them, the role position of the second virtual character may include the role position of the second virtual character within the first range, and the first range may be determined based on the current position of the first virtual character and the range of the weapon used by the first virtual character. The combat distance may be determined based on the position of the first virtual character and the role position of the second virtual character. For example, multiple distances between the first virtual character and the second virtual character may be calculated based on the current position of the first virtual character and the role positions of multiple second virtual characters, and the average of the multiple distances may be used as the combat distance. The bunker coverage rate can be determined based on the position of the virtual bunker. For example, a preset range can be determined based on the current position of the first virtual character and the role position of the second virtual character, and the total bunker coverage area and the effective combat area area within the preset range are calculated. The bunker coverage rate is determined according to the ratio of the total bunker coverage area to the effective combat area area. Among them, the occupied area of ​​each bunker can be determined based on the coordinate position of the bunker in the world coordinate system, and the sum of the areas of the bunkers falling within the preset range is calculated, and the sum of the areas is used as the total bunker coverage area. The effective combat area refers to the area where the target character can effectively attack, dodge or defend. The combat area can be pre-delineated in the map, and the area where the combat area falls within the preset range is determined as the effective combat area, and the area of ​​the effective combat area is calculated based on the position of the effective combat area in the world coordinate system.

[0110] Furthermore, the extracted features and corresponding labels can be used as training samples and the combat prediction model can be trained according to the training method mentioned above in this step, and the corresponding data can be collected in real time during the game and input into the combat prediction model, and the combat prediction model outputs combat success rate prediction information based on the input data. Exemplarily, at least one of the combat distance, bunker coverage, comparison information of the first character data and the second character data, and the role position of the second virtual character can be input into the combat prediction model, and the combat prediction model outputs a specific combat success rate and displays it in a graphical user interface, such as Fig. 9 As shown, the battle success rate "72% winning rate" can be displayed in the graphical user interface.

[0111] In some embodiments, at least one behavior label reflecting the behavior characteristics of the first virtual character can also be determined based on the historical behavior data of the first virtual character through a dynamic adjustment model. Dynamic adjustment model The dynamic adjustment model can be implemented based on a genetic algorithm or an evolutionary strategy, and its architecture is similar to the path planning model. For details, please refer to the description in the previous step 220, which will not be repeated here. The dynamic adjustment model can be obtained through the training of the historical behavior data of the first virtual character, and the behavior data of the first virtual character can be the data corresponding to the operations performed, decisions made, and interactions generated by the player in the game environment by controlling the first virtual character. Exemplary, the attack operation, the operation of using props, or the decision operation of attacking the enemy character or the strategic operation of choosing to escape in the combat state of the first virtual character controlled by the player during the game. The behavior label can reflect the behavior pattern, preference and strategy of the virtual character controlled by the player in a specific environment or situation, and the player's game style, decision-making method, interaction habits, etc. can be determined through the behavior label. For example, if the first virtual character controlled by the player is used to escaping when encountering a battle, it is confirmed that the behavior label of the first virtual character controlled by the player can be an escape type. If the first virtual character controlled by the player is used to staying and fighting when encountering a battle, it is confirmed that the behavior label of the first virtual character controlled by the player can be a combat type. The training process of the dynamic adjustment model is similar to that of the path planning model and will not be described here.

[0112] For one or more behavior tags of the first virtual character output by the dynamic adjustment model, relevant prompt information can be further adjusted based on at least one behavior tag, and the relevant prompt information can include at least one of the following information: game behavior recommendation information, movement route prompt information, warning event prompt information, and combat success rate prediction information.

[0113] In some embodiments, at least one behavior tag output by the dynamic adjustment model can be input as input data into the game task recommendation model, and the game task recommendation model outputs game behavior recommendation information for changing the execution priority of the sub-game task based on at least one behavior tag associated with the first virtual character and data in the game, such as the task attributes of the sub-game task related to the target game task, the current position of the first virtual character, the role position of the second virtual character, the map resource distribution, etc., so as to intelligently recommend sub-game tasks that are compatible with the style or preference of the first virtual character. Exemplarily, when the behavior tag associated with the first virtual character is combat-type, combat-type sub-game tasks are recommended first.

[0114] In some embodiments, at least one behavior tag output by the dynamic adjustment model can be input as input data into the direction adjustment model. The direction adjustment model outputs moving route prompt information for adjusting the direction of travel of the first virtual character based on at least one behavior tag associated with the first virtual character and data in the game, such as the current position of the first virtual character, the position of the second virtual character, the second character data corresponding to the second virtual character, the distribution of map resources, the range of the safe area, etc., so as to intelligently and flexibly adjust the direction of travel that is compatible with the style or preference of the first virtual character. Exemplarily, when the behavior tag associated with the first virtual character is an escape type, the direction of travel with fewer second virtual characters that can be encountered is preferentially recommended.

[0115] In some embodiments, at least one behavior tag output by the dynamic adjustment model can be input as input data into the combat prediction model, and the combat prediction model outputs warning event prompt information based on at least one behavior tag associated with the first virtual character and data in the game, such as real-time map data corresponding to the first game map, so as to intelligently prompt warning event prompt information that matches the style or preference of the first virtual character. Exemplarily, when the behavior tag associated with the first virtual character is combat-type, the combat success rate is improved.

[0116] In some embodiments, natural language processing can also be performed based on at least one of the following information to obtain dialogue content, and the dialogue content can be output in the form of text and / or voice: game behavior recommendation information, mobile route prompt information, warning event prompt information, and combat success rate prediction information. Among them, natural language processing can include part-of-speech tagging, entity recognition, syntactic analysis, etc. to extract semantic and contextual information. Specifically, the dialogue content can be generated by a dialogue management system. The dialogue management system can be implemented based on a finite state machine (FSM), a rule-based dialogue system, a template-based dialogue system, or a machine learning-based dialogue manager. In some embodiments, the output content of the game task recommendation model, the direction adjustment model, or the combat prediction model can be input as input data into the dialogue management system, and the dialogue management system performs natural language processing based on the input content to generate dialogue content and output it in the form of text and / or voice. Exemplarily, when the direction adjustment model determines the direction of travel, the dialogue management system can output the dialogue content based on the direction of travel, "The current priority has been changed and the optimal route has been calculated. If you want to change it, please let me know at any time." The dialogue content can be output in the form of voice, or displayed in the form of text such as Figure 6 The graphical user interface is shown.

[0117] In some embodiments, the dialogue management system can also generate dialogue content in combination with the output of the game task recommendation model and the battle prediction model. For example, when the sub-game task to be changed by the game task recommendation model is "obtain the key to safe No. 5", the dialogue management system can further generate dialogue content in combination with the warning event prompt information output by the battle prediction model, "The current status is good, and the equipment is in the top 20% level among the competitors in the same game. It is recommended to prioritize the task of 'obtaining safe No. 5'". This dialogue content can be output in voice form, or displayed in text form on a computer such as Figure 4 The graphical user interface is shown.

[0118] In some embodiments, the game data collected in real time by the embedded system or plug-in, such as the task attributes of the sub-game tasks, real-time map data, the first character data corresponding to the first virtual character, the second character data corresponding to the second virtual character, etc., can also be input into the dialogue management system. The dialogue management system generates dialogue content based on the game data collected in real time and in combination with the output of the game task recommendation model, the direction adjustment model, or the combat prediction model. Exemplarily, when the first virtual character enters the combat state, the dialogue management system generates dialogue content based on the output of the combat prediction model and the collected character position of the second virtual character and the equipment of the second virtual character: "An enemy is detected under the stairs. The enemy has a level 1 helmet and level 1 armor, and the terrain is at a disadvantage. The cover and combat position have been highlighted, and we are ready to engage the enemy." The dialogue content can be output in the form of voice, or displayed in the form of text on a computer such as a laptop. Fig. 9 The graphical user interface is shown.

[0119] This specification also provides a game information prompt system. Fig.10 is an exemplary block diagram of a game information prompt system according to some embodiments of the present specification. A graphical user interface is provided through a terminal device, and at least part of a game match scene is displayed in the graphical user interface. The game match scene includes a first virtual character participating in the game match and a second virtual character in a different game camp from the first virtual character, wherein the first virtual character is controlled by the terminal device, and the game match corresponds to a first game map. Fig.10 As shown, in some embodiments, the game information prompt system 1000 may include a determination module 1010 , a path planning module 1020 , and an information guidance module 1030 .

[0120] The determination module 1010 is used to determine a target game task and a current position of the first virtual character, wherein the target game task is used to instruct the first virtual character to complete a specified game behavior at a specified position in the game scene.

[0121] The path planning module 1020 is used to determine the path from the current position to the specified position through a path planning model and display it in a graphical user interface, wherein the path planning model is trained by the first historical game match data corresponding to the first game map.

[0122] The information guidance module 1030 is used to generate task guidance information in a graphical user interface according to the real-time map data corresponding to the first game map during the process of the first virtual character moving along the path-finding path to the designated position, so as to reduce the task difficulty of the first virtual character to complete the target game task; wherein the real-time map data includes at least one of the following data: map resource distribution, character position of the second virtual character, first character data corresponding to the first virtual character, second character data corresponding to the second virtual character, the location of the virtual bunker, and the range of the safe area; the task guidance information includes at least one of the following data: game behavior recommendation information, moving route prompt information, warning event prompt information, and combat success rate prediction information.

[0123] In some optional embodiments, the information guidance module 1030 further includes a task recommendation unit 1031, which is used to adjust the execution priority of the sub-game task based on the task attributes of the sub-game task related to the target game task and at least one of the following data in the real-time map data corresponding to the first game map through the game task recommendation model: the current position of the first virtual character, the role position of the second virtual character, and the distribution of map resources; wherein the task attributes include the task type and the task location; displaying game behavior recommendation information for changing the execution priority of the sub-game task in the graphical user interface according to the execution priority of the sub-game task; wherein the game task recommendation model is trained by the second historical game match data corresponding to the first game map, and the second historical game match data includes the task attributes of the historical sub-game task and at least one of the following data: the behavior data corresponding to the historical first virtual character, the task completion data corresponding to the historical sub-game task, the task completion time corresponding to the historical sub-game task, and the historical map data corresponding to the first game map.

[0124] In some optional embodiments, the task recommendation unit 1031 is also used to adjust the execution priority of the sub-game task based on the task attributes of the sub-game task related to the target game task and at least one of the real-time map data corresponding to the first game map through the game task recommendation model, including at least one of the following processing: determining the execution distance corresponding to the sub-game task based on the task location and the current position of the first virtual character through the game task recommendation model, and adjusting the execution priority of the sub-game task based on the execution distance; determining the safety corresponding to the sub-game task based on the character position of the second virtual character and the task location of the sub-game task through the game task recommendation model, and adjusting the execution priority of the sub-game task based on the safety; determining the estimated profit corresponding to the sub-game task based on the task location and map resource distribution of the sub-game task through the game task recommendation model, and adjusting the execution priority of the sub-game task based on the estimated profit.

[0125] In some optional embodiments, the information guidance module 1030 further includes a direction adjustment unit 1032, which is used to determine the direction of travel of the first virtual character based on at least one of the following data in the real-time map data corresponding to the first game map through a direction adjustment model: the current position of the first virtual character, the role position of the second virtual character, the second role data corresponding to the second virtual character, the map resource distribution, and the range of the safe area; according to the direction of travel of the first virtual character, the moving route prompt information for adjusting the direction of travel of the first virtual character is displayed in the graphical user interface; wherein the direction adjustment model is trained by the third historical game match data corresponding to the first game map, and the third historical game match data includes the behavior data corresponding to the historical first virtual character and at least one of the following data: the resource collection data of the historical first virtual character, the historical map data corresponding to the first game map; the character data includes at least one of health, equipment, and supply resources.

[0126] In some optional embodiments, the direction adjustment unit 1032 is also used to determine the direction of travel of the first virtual character based on at least one data in the real-time map data corresponding to the first game map through the direction adjustment model, including at least one of the following processes: in response to an instruction to change a sub-game task, determining the direction of travel from the current position of the first virtual character to the task location corresponding to the changed sub-game task through the direction adjustment model, wherein the sub-game task is related to the target game task; determining the collectible resources corresponding to the path from the current position of the first virtual character to the task location corresponding to the sub-game task being executed based on the map resource distribution through the direction adjustment model, and adjusting the direction of travel of the first virtual character according to the resource location of the collectible resources; determining the danger level information within the range corresponding to the current position of the first virtual character based on the position of the second virtual character and the second character data corresponding to the second virtual character through the direction adjustment model, and adjusting the direction of travel of the first virtual character according to the danger level information; determining the direction of travel of the first virtual character pointing to the safe area range based on the safe area range and the current position of the first virtual character through the direction adjustment model.

[0127] In some optional embodiments, the information guidance module 1030 further includes a resource management unit 1033, which is used to determine the required resources of the first virtual character based on the map resource distribution and the first character data through a resource management model; wherein the resource management model is trained by the fourth historical game match data corresponding to the first game map, the fourth historical game match data includes the historical resource collection data of the first virtual character and the historical map data corresponding to the first game map, and the required resources include at least one of the following data: food resources, combat-related resources, and resources for improving the health or ability of the virtual character.

[0128] In some optional embodiments, the direction adjustment unit 1032 is further configured to generate, in a graphical user interface, movement route prompt information for adjusting the traveling direction of the first virtual character based on the required resources of the first virtual character and the map resource distribution through a direction adjustment model.

[0129] In some optional embodiments, the information guidance module 1030 further includes a combat prediction unit 1034, which is used to generate warning event prompt information in a graphical user interface in response to the first virtual character entering a combat state through a combat prediction model based on real-time map data corresponding to the first game map, wherein the combat prediction model is trained through the fifth historical game match data corresponding to the first game map, and the fifth historical game match data includes behavior data corresponding to the historical first virtual character and at least one of the following data: resource collection data of the historical first virtual character, and historical map data corresponding to the first game map.

[0130] In some optional embodiments, the warning event prompt information includes at least one of the following information: combat position, target bunker, number of bunkers, terrain conditions, battle intensity, and second virtual character status; the battle prediction unit 1034 is also used to display warning event prompt information indicating the combat position and / or target bunker in a graphical user interface based on the current position of the first virtual character, the role position of the second virtual character, and the position of the virtual bunker through a battle prediction model; display the number of bunkers in a graphical user interface based on the position of the virtual bunker through a battle prediction model; display the terrain conditions in a graphical user interface based on the current position of the first virtual character and the role position of the second virtual character through a battle prediction model; display the battle intensity in a graphical user interface based on at least one of the following information through a battle prediction model: the role position of the second virtual character, comparison information between the first role data and the second role data, wherein the role data includes at least one of health, equipment, and supply resources; highlight the second virtual character status in a graphical user interface based on the role position and the second role data of the second virtual character through a battle prediction model, wherein the second virtual character status includes at least one of the role position of the second virtual character, the area where the second virtual character is located, the equipment status of the second virtual character, and the health status of the second virtual character.

[0131] In some optional embodiments, the combat prediction unit 1034 is also used to generate combat success rate prediction information through a combat prediction model based on at least one of the following information: combat distance, cover coverage, comparison information of the first character data and the second character data, and the role position of the second virtual character; wherein the combat distance is determined based on the current position of the first virtual character and the role position of the second virtual character; and the cover coverage is determined based on the position of the virtual cover.

[0132] In some optional embodiments, the combat prediction unit 1034 is further used to determine whether the first virtual character enters a combat state according to preset conditions; wherein, determining whether the first virtual character enters a combat state according to preset conditions includes: determining a combat distance based on a current position of the first virtual character and a character position of a second virtual character, and when the combat distance is less than a preset distance, determining that the first virtual character enters a combat state; or, when the health value in the first character data decreases, determining that the first virtual character enters a combat state; or, when the first virtual character performs a preset operation, determining that the first virtual character enters a combat state.

[0133] In some optional embodiments, the game information prompt system 1000 may further include a natural language processing module 1040, which is used to perform natural language processing based on at least one of the following information to obtain dialogue content, and output the dialogue content in the form of text and / or voice: game behavior recommendation information, movement route prompt information, warning event prompt information, combat success rate prediction information.

[0134] In some optional embodiments, the information guidance module 1030 further includes a dynamic adjustment unit 1035, which is used to determine at least one behavior label that reflects the behavior characteristics of the first virtual character based on the historical behavior data of the first virtual character through a dynamic adjustment model; adjust relevant prompt information based on at least one behavior label, and the relevant prompt information includes at least one of the following information: game behavior recommendation information, movement route prompt information, warning event prompt information, and combat success rate prediction information.

[0135] For more information about each module, see Figures 2~9 The relevant description of will not be repeated here. It should be understood that Fig.10 The system and its modules shown can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will understand that the above methods and systems can be implemented using computer executable instructions and / or control codes contained in a processor, such as a disk, CD or DVD-ROM, etc., and such codes are provided in a carrier medium or a memory of a programmable device. The system and its modules of this specification can not only be implemented by hardware circuits such as ultra-large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., but can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above hardware circuits and software (e.g., firmware).

[0136] It should be noted that the above description of the system and its modules is only for convenience of description and does not limit this specification to the scope of the embodiments. It is understandable that for those skilled in the art, after understanding the principle of the system, it is possible to arbitrarily combine the modules without deviating from this principle to form a subsystem connected to other modules. Or some modules may be split to obtain more modules or multiple units under the module. Such variations are all within the scope of the disclosure of this specification.

[0137] Some embodiments of the present specification also provide a computer program product, including computer instructions. When at least part of the computer instructions are executed by a processor, the present specification can be implemented. Figures 2~9 In some embodiments, the computer program product may only involve computer instructions, which may be carried by a storage medium or a processing device. In other embodiments, the computer program product may also be a storage medium or a processing device containing the aforementioned computer instructions. The processing device may include one or more processors and a storage medium.

[0138] In some embodiments, the processor may be a combination of one or more of the following processors: a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction set processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor (DSP), a field programmable gate array (FPGA), a programmable logic device (PLD), a programmable logic controller (PLC), a reduced instruction set computer (RISC), and a microprocessor.

[0139] In some embodiments, the storage medium may include a combination of one or more of the following: a mass storage, a removable storage, a volatile read-write memory, and a read-only memory (ROM). Exemplary mass storage may include a magnetic disk, an optical disk, a solid-state hard disk, and the like. Exemplary removable storage may include a flash disk, a floppy disk, an optical disk, a memory card, a compressed hard disk, a magnetic tape, and the like. Exemplary volatile read-write memory may include a random access memory (RAM). Exemplary random access memory may include a dynamic random access memory (DRAM), a double data rate synchronous dynamic random access memory (DDRSDRAM), a static random access memory (SRAM), a thyristor random access memory (T-RAM), and a zero capacitance memory (Z-RAM), and the like. Exemplary read-only memory may include a masked read-only memory (MROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a compact disk read-only memory (CD-ROM), and a digital multifunction hard disk read-only memory, and the like.

[0140] The beneficial effects that may be brought about by the embodiments of this specification include but are not limited to: (1) by determining the current position of the target game task and the first virtual character, and determining the path-finding path from the current position to the designated position indicated by the target game task through a path planning model, and in the process of the first virtual character moving along the path-finding path to the designated position, task guidance information is generated in a graphical user interface according to the real-time map data corresponding to the first game map, so as to reduce the task difficulty of the first virtual character to complete the target game task, lower the game threshold, enable players to quickly understand the situation in the game game scene, help players enter into a deep play state such as conceiving battle strategies and coordinated operations according to the task guidance information, and enhance the player's gaming experience; (2) through a game task recommendation model, the execution priority of the sub-game task is displayed in a graphical user interface based on the real-time map data corresponding to the first game map. (3) through the direction adjustment model, based on the real-time map data corresponding to the first game map, the moving route prompt information for adjusting the direction of travel of the first virtual character is displayed in the graphical user interface, so as to flexibly indicate the direction of travel of the first virtual character controlled by the player in the game according to the real-time map data, and assist the player to quickly familiarize himself with the game map and understand the current game environment; (4) through the battle prediction model, based on the real-time map data corresponding to the first game map, the warning event prompt information is displayed in the graphical user interface, so as to assist the player to understand the current game environment, intelligently provide relevant suggestions to the player, lower the learning threshold of the player, and improve the player's game experience. At the same time, it also provides new gameplay and enhances the strategic depth.

[0141] It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects that may be produced may be any one or a combination of the above, or any other possible beneficial effects.

[0142] The basic concepts have been described above. Obviously, for those skilled in the art, the above detailed disclosure is only for example and does not constitute a limitation of this specification. Although not explicitly stated here, those skilled in the art may make various modifications, improvements and corrections to this specification. Such modifications, improvements and corrections are taught in this specification, so such modifications, improvements and corrections still belong to the spirit and scope of the exemplary embodiments of this specification.

Claims

1. A game information prompting method, characterized in that: A graphical user interface is provided through a terminal device, wherein at least a portion of a game scene is displayed in the graphical user interface, wherein the game scene includes a first virtual character participating in the game and a second virtual character in a different game camp from the first virtual character, wherein the first virtual character is controlled through the terminal device, and the game corresponds to a first game map, and the method includes: Determining a target game task and a current position of the first virtual character, wherein the target game task is used to instruct the first virtual character to complete a specified game behavior at a specified position in the game scene; Determining a path from the current position to the designated position through a path planning model, and displaying the path in the graphical user interface, wherein the path planning model is trained by first historical game match data corresponding to the first game map; During the process of the first virtual character moving along the path-finding path to the designated position, task guidance information is generated in the graphical user interface according to the real-time map data corresponding to the first game map, so as to reduce the task difficulty of the first virtual character in completing the target game task; wherein the real-time map data includes at least one of the following data: map resource distribution, character position of the second virtual character, first character data corresponding to the first virtual character, second character data corresponding to the second virtual character, the position of the virtual bunker, and the range of the safe area; the task guidance information includes at least one of the following data: game behavior recommendation information, movement route prompt information, warning event prompt information, and combat success rate prediction information.

2. The method according to claim 1, characterized in that The path planning model is trained by using historical path data in the first historical game match data corresponding to the first game map and at least one of the following data: task completion data corresponding to the historical path data, casualty data corresponding to the historical path data, task completion time corresponding to the historical path data, and revenue data corresponding to the historical path data.

3. The method according to claim 1, characterized in that Generating task guidance information in the graphical user interface according to the real-time map data corresponding to the first game map includes: The execution priority of the sub-game task is adjusted based on the task attributes of the sub-game task related to the target game task and at least one of the following data in the real-time map data corresponding to the first game map through the game task recommendation model: the current position of the first virtual character, the role position of the second virtual character, and the map resource distribution; wherein the task attributes include task type and task location; Displaying game behavior recommendation information for changing the execution priority of the sub-game task in the graphical user interface according to the execution priority of the sub-game task; Among them, the game task recommendation model is trained by the second historical game match data corresponding to the first game map, and the second historical game match data includes task attributes of historical sub-game tasks and at least one of the following data: behavior data corresponding to the historical first virtual character, whether the task corresponding to the historical sub-game task is completed, the task completion time corresponding to the historical sub-game task, and historical map data corresponding to the first game map.

4. The method according to claim 3, characterized in that Adjusting the execution priority of the sub-game task based on the task attribute of the sub-game task related to the target game task and at least one of the real-time map data corresponding to the first game map through the game task recommendation model includes at least one of the following processes: Determining the execution distance corresponding to the sub-game task based on the task location and the current position of the first virtual character by the game task recommendation model, and adjusting the execution priority of the sub-game task based on the execution distance; Determining the safety corresponding to the sub-game task based on the role position of the second virtual character and the task location of the sub-game task through the game task recommendation model, and adjusting the execution priority of the sub-game task based on the safety; The game task recommendation model determines an estimated revenue corresponding to the sub-game task based on the task location of the sub-game task and the map resource distribution, and adjusts the execution priority of the sub-game task based on the estimated revenue.

5. The method according to claim 1, characterized in that Generating task guidance information in the graphical user interface according to the real-time map data corresponding to the first game map includes: Determining the direction of travel of the first virtual character based on at least one of the following data in the real-time map data corresponding to the first game map through a direction adjustment model: the current position of the first virtual character, the role position of the second virtual character, the second role data corresponding to the second virtual character, the map resource distribution, and the range of the safe area; Displaying, in the graphical user interface, movement route prompt information for adjusting the movement direction of the first virtual character according to the movement direction of the first virtual character; Among them, the direction adjustment model is trained by the third historical game match data corresponding to the first game map, and the third historical game match data includes behavior data corresponding to the historical first virtual character and at least one of the following data: resource collection data of the historical first virtual character, historical map data corresponding to the first game map; the character data includes at least one of health, equipment, and supply resources.

6. The method according to claim 5, characterized in that Determining the traveling direction of the first virtual character based on at least one data in the real-time map data corresponding to the first game map by using the direction adjustment model includes at least one of the following processes: In response to an instruction to change a sub-game task, determining a travel direction from a current position of the first virtual character to a task location corresponding to the changed sub-game task by using the direction adjustment model, wherein the sub-game task is related to the target game task; Determining, by the direction adjustment model and based on the map resource distribution, the collectible resources corresponding to the path from the current position of the first virtual character to the task location corresponding to the sub-game task being executed, and adjusting the travel direction of the first virtual character according to the resource position of the collectible resources; determining, by the direction adjustment model, danger information within a range corresponding to the current position of the first virtual character based on the position of the second virtual character and second character data corresponding to the second virtual character, and adjusting the traveling direction of the first virtual character according to the danger information; The direction of travel of the first virtual character pointing to the safety area range is determined by the direction adjustment model based on the safety area range and the current position of the first virtual character.

7. The method according to claim 5, characterized in that The moving route prompt information also includes at least one of the following information: danger information related to the path finding path, estimated benefit information related to the path finding path, estimated time related to the path finding path, and alternative shelters related to the path finding path; The generating of task guidance information in the graphical user interface according to the real-time map data corresponding to the first game map further comprises at least one of the following processes: displaying danger information related to the pathfinding path in the graphical user interface based on the character position of the second virtual character; displaying estimated revenue information related to the pathfinding path in the graphical user interface based on the map resource distribution; displaying in the graphical user interface an estimated time to reach the designated location based on the current location of the first virtual character, the designated location, and a moving speed of the first virtual character; Based on the position of the virtual cover, alternative cover information related to the path finding path is displayed in the graphical user interface and the alternative cover is highlighted.

8. The method according to claim 5, characterized in that The generating task guidance information in the graphical user interface according to the real-time map data corresponding to the first game map further comprises: Determining the required resources of the first virtual character based on the map resource distribution and the first character data through a resource management model; wherein the resource management model is trained through fourth historical game match data corresponding to the first game map, the fourth historical game match data includes resource collection data of the historical first virtual character and historical map data corresponding to the first game map, and the required resources include at least one of the following data: food resources, combat-related resources, and resources for improving the health or ability of the virtual character; The moving route prompt information for adjusting the traveling direction of the first virtual character is generated in the graphical user interface through the direction adjustment model based on the required resources of the first virtual character and the map resource distribution.

9. The method according to claim 1, characterized in that: The generating task guidance information in the graphical user interface according to the real-time map data corresponding to the first game map further comprises: In response to the first virtual character entering a combat state, a warning event prompt message is generated in the graphical user interface through a combat prediction model based on real-time map data corresponding to the first game map, wherein the combat prediction model is trained through fifth historical game match data corresponding to the first game map, and the fifth historical game match data includes behavior data corresponding to the historical first virtual character and at least one of the following data: resource collection data of the historical first virtual character and historical map data corresponding to the first game map.

10. The method according to claim 9, characterized in that The warning event prompt information includes at least one of the following information: combat position, target bunker, number of bunkers, terrain conditions, intensity of battle, and status of the second virtual character; The generating of warning event prompt information in the graphical user interface based on the real-time map data corresponding to the first game map by using the battle prediction model includes at least one of the following processes: Displaying warning event prompt information indicating the combat position and / or the target bunker in the graphical user interface based on the current position of the first virtual character, the role position of the second virtual character, and the position of the virtual bunker by the combat prediction model; Displaying the number of bunkers in the graphical user interface based on the positions of the virtual bunkers by the combat prediction model; Displaying the terrain condition in the graphical user interface based on the current position of the first virtual character and the character position of the second virtual character by using the battle prediction model; Displaying the intensity of the battle in the graphical user interface by the battle prediction model based on at least one of the following information: a character position of the second virtual character, and comparison information between the first character data and the second character data, wherein the character data includes at least one of health, equipment, and supply resources; The second virtual character status is highlighted in the graphical user interface based on the character position of the second virtual character and the second character data through the combat prediction model, and the second virtual character status includes at least one of the character position of the second virtual character, the area where the second virtual character is located, the equipment status of the second virtual character, and the health status of the second virtual character.

11. The method according to claim 9, characterized in that The generating task guidance information in the graphical user interface according to the real-time map data corresponding to the first game map further comprises: The combat prediction model generates combat success rate prediction information based on at least one of the following information: combat distance, cover coverage, comparison information of the first character data and the second character data, and the role position of the second virtual character; wherein the combat distance is determined based on the current position of the first virtual character and the role position of the second virtual character; and the cover coverage is determined based on the position of the virtual cover.

12. The method according to claim 11, characterized in that The character position of the second virtual character includes the character position of the second virtual character within a first range; the first range is determined based on the current position of the first virtual character and the range of a weapon used by the first virtual character.

13. The method according to claim 9, characterized in that Also includes: Determining whether the first virtual character enters a combat state according to preset conditions; Wherein, judging whether the first virtual character enters a combat state according to a preset condition comprises: determining a combat distance based on the current position of the first virtual character and the position of the second virtual character, and determining that the first virtual character enters a combat state when the combat distance is less than a preset distance; or When the health value in the first character data decreases, it is determined that the first virtual character enters a combat state; or, When the first virtual character performs a preset operation, it is determined that the first virtual character enters a combat state.

14. The method according to claim 1, characterized in that Also includes: The conversation content is obtained by performing natural language processing based on at least one of the following information, and the conversation content is output in the form of text and / or voice: the game behavior recommendation information, the movement route prompt information, the warning event prompt information, and the combat success rate prediction information.

15. The method according to claim 1, characterized in that Also includes: Determining at least one behavior label reflecting the behavior characteristics of the first virtual character based on the historical behavior data of the first virtual character by dynamically adjusting the model; The relevant prompt information is adjusted based on the at least one behavior tag, wherein the relevant prompt information includes at least one of the following information: the game behavior recommendation information, the movement route prompt information, the warning event prompt information, and the combat success rate prediction information.

16. A game information prompt system, characterized in that: A graphical user interface is provided through a terminal device, wherein at least a part of a game scene is displayed in the graphical user interface, wherein the game scene includes a first virtual character participating in the game and a second virtual character in a different game camp from the first virtual character, wherein the first virtual character is controlled through the terminal device, and the game corresponds to a first game map, and the system includes: A determination module, used to determine a target game task and a current position of the first virtual character, wherein the target game task is used to instruct the first virtual character to complete a specified game behavior at a specified position in the game scene; a path planning module, configured to determine a path from the current position to the specified position through a path planning model, and display the path in the graphical user interface, wherein the path planning model is trained by first historical game match data corresponding to the first game map; An information guidance module is used to generate task guidance information in the graphical user interface according to the real-time map data corresponding to the first game map during the process of the first virtual character moving along the path-finding path to the designated position, so as to reduce the task difficulty of the first virtual character to complete the target game task; wherein the real-time map data includes at least one of the following data: map resource distribution, character position of the second virtual character, first character data corresponding to the first virtual character, second character data corresponding to the second virtual character, the position of the virtual bunker, and the range of the safe area; the task guidance information includes at least one of the following data: game behavior recommendation information, movement route prompt information, warning event prompt information, and combat success rate prediction information.

17. A computer program product, characterized in that The invention comprises a computer program, and when at least a part of the computer program is executed by a processor, the method according to any one of claims 1 to 15 can be implemented.

Images