Game data processing method and device, equipment, storage medium and program product

Abstract

The invention provides a game data processing method and device, equipment, a storage medium and a program product. The method comprises the steps of obtaining first graphic data of various first blocks currently displayed in a man-machine interaction interface of a game and first state information of each first block; calling a deep learning model to perform operation prediction processing according to the first graphic data and the first state information to obtain an operation prediction result; generating prompt information according to the operation prediction result and game configuration information; the prompt information is displayed in a pre-configuration area in the human-computer interaction interface, the prompt information is used for representing the positions of a preset number of second image blocks, the second image blocks are first image blocks of the same type in the to-be-selected state and used for executing selection operation, and when the preset number of second image blocks are selected, the first image blocks in the to-be-selected state are selected. And hiding the preset number of second blocks. Through the method and the device, the information for prompting the operation can be efficiently generated in real time.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, device, equipment, storage medium and program product for processing game data. Background Art

[0002] With the advancement of online game technology in recent years, various online games and game play methods have emerged in an endless stream. During the game process, for example: graphic elimination games, such games often contain a variety of graphic elements, props and preset rules. Players move graphics to meet the requirements of passing the level. Different levels correspond to different strategies. When players cannot think about how to perform the next operation in the current situation, it is difficult to continue the game and lose interest in the game. The relevant technology generates prompts based on predefined fixed rules, lacks the ability to identify and dynamically adjust the personalized needs of players, and may interfere with the player's gaming experience and the overall fluency of the game.

[0003] In the related technology, there is no more real-time and efficient method to generate prompt information. Summary of the invention

[0004] The embodiments of the present application provide a method, apparatus, device, storage medium and program product for processing game data, which can generate information for prompting operations in real time and efficiently.

[0005] The technical solution of the embodiment of the present application is implemented as follows:

[0006] The present application provides a method for processing game data, the method comprising:

[0007] Acquire first graphic data of multiple first blocks currently displayed in the human-computer interaction interface of the game and first state information of each of the first blocks;

[0008] There are multiple first blocks of the same type, and the graphics corresponding to the first blocks of the same type are the same, and the state information includes: a waiting state and a prohibited selection state;

[0009] Calling a deep learning model to perform operation prediction processing according to the first graphic data and the first state information to obtain an operation prediction result;

[0010] Generate prompt information according to the operation prediction result and game configuration information;

[0011] The prompt information is displayed in a preconfigured area in the human-computer interaction interface, wherein the prompt information is used to represent the position of a preset number of second tiles, the second tiles are the same type of first tiles in the to-be-selected state for performing a selection operation, and when the preset number of second tiles are selected, the preset number of second tiles are hidden.

[0012] The present application provides a device for processing game data, including:

[0013] A data acquisition module, used to acquire first graphic data of multiple first blocks currently displayed in the human-computer interaction interface of the game and first state information of each first block; wherein the number of the same first blocks is multiple, the graphics corresponding to the same first blocks are the same, and the state information includes: a waiting state and a prohibited selection state;

[0014] An operation prediction module is used to call a deep learning model to perform operation prediction processing according to the first graphic data and the first state information to obtain an operation prediction result; generate prompt information according to the operation prediction result and game configuration information; and display the prompt information in a preconfigured area in the human-computer interaction interface, wherein the prompt information is used to represent the position of a preset number of second tiles, the second tiles are the same type of first tiles in the to-be-selected state for performing a selection operation, and when the preset number of second tiles are selected, the preset number of second tiles are hidden.

[0015] An embodiment of the present application provides an electronic device, the electronic device comprising:

[0016] A memory for storing computer executable instructions or computer programs;

[0017] The processor is used to implement the method for processing game data provided in the embodiment of the present application when executing the computer executable instructions or computer programs stored in the memory.

[0018] An embodiment of the present application provides a computer-readable storage medium storing a computer program or computer-executable instructions for implementing a method for processing game data provided by an embodiment of the present application when executed by a processor.

[0019] An embodiment of the present application provides a computer program product, including a computer program or computer executable instructions. When the computer program or computer executable instructions are executed by a processor, the game data processing method provided by the embodiment of the present application is implemented.

[0020] The embodiments of the present application have the following beneficial effects:

[0021] By acquiring the first graphic data and first state information of the various first blocks currently displayed in the human-computer interaction interface of the game, the deep learning model is called based on the first graphic data and the first state information to perform operation prediction processing, and the operation prediction results are obtained. By accurately predicting the possible operations of the user, the strategy and fun of the game are increased. Prompt information is generated based on the operation prediction results and the game configuration information and displayed to the player. The player thinks and chooses according to the prompt information, which improves the selection operation efficiency and the fluency of the game, and helps to improve the overall quality of the game and enhance the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is a schematic diagram of an application mode of the method for processing game data provided in an embodiment of the present application;

[0023] Figure 2 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0024] Figure 3A is a first flow chart of a method for processing game data provided by an embodiment of the present application;

[0025] Figure 3B is a second flow chart of the method for processing game data provided in an embodiment of the present application;

[0026] Figure 3C is a third flow chart of the method for processing game data provided in an embodiment of the present application;

[0027] Figure 4 is a fourth flow chart of the method for processing game data provided in an embodiment of the present application;

[0028] Figure 5 It is a schematic diagram of the structure of the generative pre-training model provided in the embodiment of the present application;

[0029] Figure 6 It is a schematic diagram of the human-computer interaction interface provided in an embodiment of the present application.

[0030] It should be pointed out that the above-mentioned "first" and "second" are only used to distinguish different solutions, and do not represent the degree of superiority or inferiority of the solutions or the priority in the implementation process. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0032] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0033] In the following description, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein.

[0034] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0035] Unless otherwise defined, all technical and scientific terms used in the embodiments of the present application have the same meanings as those commonly understood by those skilled in the art. The terms used in the embodiments of the present application are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0036] The collection and processing of relevant data (for example, user game information and interface information of the human-computer interaction interface) in the embodiments of this application should be strictly in accordance with the requirements of relevant laws and regulations when applied in examples, and the informed consent or separate consent of the personal information subject should be obtained, and subsequent data use and processing should be carried out within the scope of authorization of laws and regulations and the personal information subject.

[0037] Before further describing the embodiments of the present application in detail, the nouns and terms involved in the embodiments of the present application are explained. The nouns and terms involved in the embodiments of the present application are subject to the following interpretations.

[0038] 1) Generative Pre-trained Transformer (GPT): It is a natural language processing model based on deep learning. It is trained with massive pre-training data. The generative pre-training model performs unsupervised learning on large-scale data sets to learn the statistical characteristics and inherent laws of the data. After pre-training, the generative pre-training model can be used to generate new data samples that are statistically similar to the training data. The sample generation process is autoregressive and predicts the next part based on the previous part of the input.

[0039] 2) Deep Reinforcement Learning (DRL): A machine learning method that learns the optimal strategy through the interaction between the agent and the environment to maximize the cumulative reward.

[0040] 3) Game Mechanics: These are the basic rules and systems that make up the gameplay, defining the actions and possible interactions that players can perform in the game. Game mechanics can include scoring rules, time limits, resource management, character abilities, upgrade systems, etc. These mechanisms drive the game's progress and the player's decisions.

[0041] 4) Core Gameplay Loop: refers to a series of key activities and interactions that players repeatedly perform in the game, which constitute the main experience of the game and the continuous motivation of the players. The core gameplay framework usually includes a set of basic operations or decisions, such as exploration, resource collection, and task completion, which are looped through specific rules and feedback mechanisms. This framework not only defines the basic rhythm and challenge level of the game, but also improves the entertainment and sustainability of the game.

[0042] 5) User Interface (UI): It is the bridge for users to interact with the system, including the design and layout of the interface.

[0043] In the related art, prompts are generated based on predefined fixed rules, which lack the ability to identify and dynamically adjust players' personalized needs. It is impossible to provide personalized prompts based on the real-time status of the game and the players' personal habits, which may interfere with the players' gaming experience and the overall smoothness of the game.

[0044] Embodiments of the present application provide a method for processing game data, a device for processing game data, an electronic device, a computer-readable storage medium, and a computer program product, which can efficiently generate information for prompting operations in real time.

[0045] The following describes an exemplary application of the electronic device provided in the embodiments of the present application. The device provided in the embodiments of the present application can be implemented as various types of terminals such as a laptop computer, a tablet computer, a desktop computer, a set-top box, a smart phone, a smart speaker, a smart watch, a smart TV, and a vehicle-mounted terminal, and can also be implemented as a server. The following describes an exemplary application when the device is implemented as a terminal or a server.

[0046] See also Figure 1 , Figure 1 Schematic diagram of the application mode of the method for processing game data provided in the embodiment of the present application, for implementing a processing application supporting game data, for example, Figure 1 The system involves a server 200, a network 300, a terminal device 400 and a database 500. The terminal device 400 is connected to the server 200 via the network 300. The network 300 can be a wide area network or a local area network, or a combination of the two.

[0047] In some embodiments, the user may be a technician in the relevant field or a player, the server 200 is a server for executing a method for processing game data, the terminal device 400 is a terminal operated by the user, the terminal device 400 is pre-configured with an application for displaying a virtual scene of the game, and the database 500 stores user history information and game configuration information.

[0048] For example, a user triggers terminal device 400 to experience a game, and terminal device 400 sends information about the user's current game situation to server 200 via network 300. Server 200 reads user history information from database 500, predicts the player's operation through a deep learning model, and obtains an operation prediction result with the highest probability of success. Strategy prompt information is generated based on the operation prediction result and game configuration information. Server 200 sends the strategy prompt information to terminal device 400 via network 300, and the user continues to operate the game on terminal device 400 based on the strategy prompt information, thereby improving user experience.

[0049] The embodiments of the present application can be implemented through database technology. In short, a database can be regarded as an electronic file cabinet where electronic files are stored. Users can add, query, update, delete, etc. data in the files. The so-called "database" is a collection of data that is stored together in a certain way, can be shared with multiple users, has as little redundancy as possible, and is independent of the application program.

[0050] A database management system (DBMS) is a computer software system designed for managing databases. It generally has basic functions such as storage, retrieval, security, and backup. Database management systems can be classified according to the database model they support, such as relational, XML (Extensible Markup Language); or according to the type of computer they support, such as server clusters, mobile phones; or according to the query language used, such as Structured Query Language (SQL), XQuery; or according to performance focus, such as maximum scale, maximum operating speed; or other classification methods. Regardless of the classification method used, some DBMS can cross categories, for example, supporting multiple query languages ​​at the same time.

[0051] See also Figure 2 , Figure 2 is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. The electronic device may be Figure 1 Server 200 in Figure 2 The server 200 shown includes: at least one processor 410, a memory 450, and at least one network interface 420. The various components in the server 200 are coupled together via a bus system 440. It is understood that the bus system 440 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 440 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, the bus system 440 is not described in detail. Figure 2 Various buses are labeled as bus system 440 .

[0052] The processor 410 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., where the general-purpose processor can be a microprocessor or any conventional processor, etc.

[0053] The memory 450 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid-state memory, hard disk drives, optical disk drives, etc. The memory 450 may optionally include one or more storage devices that are physically remote from the processor 410.

[0054] The memory 450 includes a volatile memory or a nonvolatile memory, and may also include both volatile and nonvolatile memories. The nonvolatile memory may be a read-only memory (ROM), and the volatile memory may be a random access memory (RAM). The memory 450 described in the embodiments of the present application is intended to include any suitable type of memory.

[0055] In some embodiments, memory 450 can store data to support various operations, examples of which include programs, modules, and data structures, or a subset or superset thereof, as exemplarily described below.

[0056] Operating system 451, including system programs for processing various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks;

[0057] The network communication module 452 is used to reach other electronic devices via one or more (wired or wireless) network interfaces 420. Exemplary network interfaces 420 include: Bluetooth, Wireless Authentication (WiFi), and Universal Serial Bus (USB).

[0058] In some embodiments, the device provided in the embodiments of the present application can be implemented in software. Figure 2 The processing 455 of the game data stored in the memory 450 is shown, which can be software in the form of programs and plug-ins, including the following software modules: a data acquisition module 4551 and an operation prediction module 4552. These modules are logical, so they can be arbitrarily combined or further split according to the functions implemented. The functions of each module will be explained below.

[0059] In some embodiments, the terminal or server can implement the processing method of question and answer data provided by the embodiment of the present application by running various computer executable instructions or computer programs. For example, computer executable instructions can be commands, machine instructions or software instructions at the microprogram level. The computer program can be a native program or software module in the operating system; it can be a local (Native) application (APPlicat ion, APP), that is, a program that needs to be installed in the operating system to run; it can also be a small program that can be embedded in any APP, that is, a program that can be run only by downloading it to a browser environment. In short, the above-mentioned computer executable instructions can be instructions in any form, and the above-mentioned computer program can be an application, module or plug-in in any form.

[0060] The method for processing game data provided by the embodiment of the present application will be explained in combination with the exemplary application and implementation of the terminal provided by the embodiment of the present application.

[0061] The following describes the method for processing game data provided by the embodiment of the present application. As mentioned above, the electronic device for implementing the method for processing question and answer data in the embodiment of the present application may be a terminal, a server, or a combination of the two. Therefore, the execution subject of each step will not be repeatedly described below.

[0062] See also Figure 3A , Figure 3A is a first flow chart of the method for processing game data provided by the embodiment of the present application, which will be combined with Figure 3A The steps shown are explained.

[0063] In step 301, first graphic data of multiple first blocks currently displayed in a human-computer interaction interface of a game and first state information of each first block are obtained.

[0064] Here, there are multiple first blocks of the same type, the graphics corresponding to the first blocks of the same type are the same, and the state information includes: a waiting-to-be-selected state and a prohibited-selection state.

[0065] For example, the information currently displayed in the human-computer interaction interface of the game is obtained. In the embodiment of the present application, the game is a graphic elimination game. The graphic elimination game is a puzzle game. During the game, the player mainly matches a certain amount of the same game elements, such as fruits, gems, animal portraits, etc., to eliminate them to win the game. In the graphic elimination game, there are multiple types of first tiles, and there are multiple tiles of the same type, and the graphics of each type of tile are the same. The user selects the first tile to eliminate the first tile in the human-computer interaction interface. The elimination of the first tile can be used to obtain the corresponding clearance score or reward mechanism. The state of each first tile includes a waiting state and a prohibited selection state. The state information of the first tile is updated with the user's operation on the first tile. The first graphic data includes the position information and graphic type information of the first graphic.

[0066] In step 302, a deep learning model is called to perform operation prediction processing according to the first graphic data and the first state information to obtain an operation prediction result.

[0067] For example, in an embodiment of the present application, the deep learning model is a generative pre-trained model. The generative pre-trained model is a natural language processing model based on deep learning, which is used to generate new data samples. These samples are statistically similar to the training data. The sample generation process is autoregressive, and the next part is predicted based on the previous part of the input. The deep learning model is called according to the first graphic data and the first state information to perform operation prediction processing. The generative pre-trained model predicts the optimal solution strategy for the information currently displayed in the human-computer interaction interface, and the predicted solution strategy is used as the prediction result of the user's next operation. The optimal solution strategy is the operation performed by the user that the generative pre-trained model predicts to maximize the probability of the user passing the level.

[0068] In some embodiments, based on the first graphic data and the first state information, a deep learning model is called to perform operation prediction processing, see Figure 3B , Figure 3B is a second flow chart of the method for processing game data provided in an embodiment of the present application; Figure 3A Step 302 in the above example can be accomplished by using a deep learning model and Figure 3B Steps 3021 to 3022 in are implemented as described in detail below.

[0069] In step 3021, feature extraction processing is performed on the first graphic data and the first state information to obtain current game features.

[0070] For example, feature extraction is performed on the first graphic data and the first state information. The feature extraction is implemented by a deep learning model. After receiving the input first graphic data and the first state information, the deep learning model converts the discrete input data into a continuous vector representation for feature extraction to obtain the current game features.

[0071] In some embodiments, see Figure 5 , Figure 5 501 is a schematic diagram of the structure of the generative pre-training model provided in the embodiment of the present application; in the embodiment of the present application, the deep learning model is a generative pre-training model, and the generative pre-training model 501 is composed of an input layer 5011, an embedding layer 5012, a generative pre-training layer 5013, a reinforcement learning combination layer 5014, a decoding layer 5015, and an output layer 5016. The first graphic data and the first state information are subjected to feature extraction processing, and the received data is standardized through the input layer 5011, and is passed to the embedding layer 5012 to convert the discrete data into a continuous vector representation, and then the multi-layer attention mechanism of the generative pre-training layer 5013 is used to realize the extraction of the current game features.

[0072] In some embodiments, step 3021 can be implemented by the following method: converting the first graphic data and the first state information to obtain a first embedded vector representation, wherein the first embedded vector representation is represented as a sequence of multiple feature vectors; normalizing the first embedded vector representation to obtain a second embedded vector representation; performing multiple feature extraction processes on the second embedded vector representation to obtain current game features, wherein the input of each feature extraction process is the output of the previous feature extraction process.

[0073] For example, the input layer 5011 receives the first graphic data and the first state information, and converts the first graphic data and the first state information through the embedding layer 5012 to capture the semantic information in the discrete data, convert the discrete data into a continuous vector representation, and extract the continuous vector features through the multi-layer attention mechanism in the generative pre-training layer 5013 to obtain a first embedded vector representation, which is a sequence composed of multiple feature vectors corresponding to the input sequence. The first embedded vector representation is normalized, the mean and standard deviation of the first embedded vector are calculated, and each feature vector is independently standardized so that each vector has a unit variance and a zero mean. For each first embedded vector, the layer normalization process applies a learnable parameter to obtain a second embedded vector representation. The second embedded vector representation is subjected to multiple feature extraction processes, and the output of the previous layer is superimposed on the output of the subsequent layer through identity mapping. The dimension of the output is adjusted through linear transformation to obtain the current game features.

[0074] Continue to see Figure 3BIn step 3022, operation strategy prediction processing is performed based on the current game characteristics to obtain operation prediction results.

[0075] For example, operation strategy prediction processing is performed based on current game features. The operation strategy prediction processing includes multiple selection operations. The deep learning model evaluates the priority of different options through previous user operations in similar human-computer interaction interfaces to obtain operation prediction results.

[0076] In some embodiments, the prediction processing of the operation strategy is implemented through a deep learning model, and the strategy is optimized by calling the reinforcement learning layer 5014 based on the level data features. Through the deep learning algorithm, the strategy is adjusted according to the historical operation information and clearance rate of the player in the historical situation, and the priority of different operation options is evaluated. The processed features are decoded by the decoding layer 5015 to generate the operation prediction results, and the prediction results of the situation changes are output through the fully connected layer. The output layer 5016 outputs the prediction results of the situation changes and the optimal strategy through an activation function (e.g., softmax).

[0077] In some embodiments, step 3022 can be implemented by the following method: performing prediction processing based on the current game characteristics to obtain different types of third tiles in the first tile to be selected; for each third tile, performing multiple prediction processing based on the current game characteristics and the same third tiles, until a preset number of the same third tiles are obtained, wherein the input of each prediction processing includes: the same third tiles that have been predicted, and the current game characteristics updated based on the predicted third tiles of the same type; selecting one from each preset number of third tiles as the second tile; using the predicted order of the second tiles and the second graphic data of each second tile as the operation prediction result, wherein the predicted order is used as the selection order of the second tiles.

[0078] For example, a prediction process is performed based on the current game feature to obtain different types of third tiles in the first tile to be selected. When the player removes the first tile, the current game feature will be updated based on the player's operation. After the removal, the previously blocked third tiles of different types will be displayed. The updated third tiles are predicted multiple times until a preset number of the same third images are obtained, which are used to update the current game feature based on the preset number of third tiles. One of the preset number of third tiles is selected as the second tile, and the predicted selection order of the second tile and the corresponding graphic information are used as the prediction result of the operation.

[0079] In some embodiments, one type is selected as the second tile from each preset number of third tiles in the following manner: for each third tile, the game success rate of a preset number of the same third tiles is predicted based on historical game operation results, wherein the game winning rate represents the success probability of the current game after the preset number of the same third tiles are hidden; and the preset number of the same third tiles with the highest game success probability are selected as the second tile.

[0080] For example, the number of third tiles includes multiple ones, and the game success rate of predicting a preset number of the same third tiles is predicted based on the game operation results of similar situations in the user's history. The game winning rate represents the success probability of the current game after the preset number of the same third tiles are hidden. The third tile with the highest success probability of the current game is selected as the second tile, and the selection order and image information corresponding to the second tile are used as prompts. Predicting the game success rate of a preset number of the same third tiles can be achieved through a machine learning model.

[0081] Continue to see Figure 3A In step 303, prompt information is generated according to the operation prediction result and the game configuration information.

[0082] For example, the game configuration information is the game mechanism. The game mechanism is the basic principle that constitutes the gameplay, which defines the operations that players can perform and the possible ways of interaction in the game. Prompt information for the human-computer interaction interface of the current game is generated based on the operation prediction results and the game configuration information.

[0083] In some embodiments, step 303 can be implemented by the following method: using a preconfigured icon as a prompt icon for each second tile; performing text generation processing based on the second graphic data of each second tile and the game configuration information according to the selection order to obtain a prompt text corresponding to each second tile; marking the prompt text of each second tile and the display order of the prompt icon according to the selection order; and using the marked prompt text and prompt icon as prompt information.

[0084] For example, the preconfigured icon is a pre-set icon that is different from the second tile, for example: a frame line is drawn outside the second tile, the border color of the second tile is modified, or the display effect of the second tile is modified. A prompt mark is made for each second icon based on the preconfigured icon, and text is generated based on the graphic data of the second tile and the game configuration information according to the selection order predicted by the second tile, and the predicted optimal solution strategy is used as the corresponding prompt text. The prompt text and the prompt icon are marked in order according to the selection order, and the prompt text and the prompt icon are displayed to the player as prompt information according to the marked information.

[0085] In step 304, prompt information is displayed in a preconfigured area in the human-computer interaction interface.

[0086] Here, the prompt information is used to indicate the positions of a preset number of second tiles, where the second tiles are the same type of first tiles in a waiting state for performing a selection operation, and when the preset number of second tiles are selected, the preset number of second tiles are hidden.

[0087] For example, the human-computer interaction interface includes design and layout. The preconfigured area is a location pre-designed for displaying prompt information. Prompt information can be intuitively displayed to players in the preconfigured area. The prompt information includes but is not limited to presentation in the form of icons, text or sound. Through the prompt information, players can clearly see the current game status analysis, recommended next operations and potential game strategies, thereby improving operating efficiency and gaming experience.

[0088] In some embodiments, see Figure 6 , Figure 6 is a schematic diagram of a human-computer interaction interface provided in an embodiment of the present application, Figure 6 The figure shows a screen of the current game in the human-computer interaction interface 601, and a prompt message 603 is displayed in the graphical interface to give the user a prompt for the next operation. In response to triggering the dynamic prompt icon 602, the prompt content of the prompt message 603 is displayed in the preconfigured area.

[0089] In some embodiments, see Figure 3C , Figure 3C is a third flow chart of the method for processing game data provided by the embodiment of the present application; before step 301, Figure 3C Steps 3011 to 3013 in the process implement the training of the deep learning model, which is described in detail below.

[0090] In step 3011, historical game data is obtained.

[0091] Here, the historical game data includes: first graphic data of the first tile in each game, selection operations for each first tile, and game results of each game.

[0092] For example, based on the premise that the core gameplay framework of the game remains unchanged, the game levels are automatically operated through self-testing and editors, a large number of game levels are simulated, the initial data of all situation changes during the simulation process are collected, and simple tests are performed to filter out the steps that directly lead to failure, and historical game data is obtained. The core gameplay framework of the game refers to a series of key activities and interactions that players repeatedly perform in the game. These activities constitute the main experience of the game and the continuous motivation of the players. The core gameplay framework usually includes a set of basic operations or decisions, which are cyclically performed through specific rules and feedback mechanisms. The data obtained from obtaining each simulated level includes: the first graphic data of each game (the location, status, and type of graphics), the selection operation for each first tile (operations such as swapping, eliminating, and moving graphics), the game results of each game, and the game panel status information (graphic arrangement status, player score, level completion conditions, and current status progress).

[0093] In step 3012, the historical game data is denoised to obtain sample game data.

[0094] For example, historical game data is preprocessed to remove noise and irrelevant information in the level data. Noise is random errors or outliers in the data. The historical game data is converted into a format suitable for model processing as sample game data for the model.

[0095] In step 3013, the deep learning model is trained based on the sample game data to obtain a trained deep learning model.

[0096] For example, the deep learning model is a generative pre-training model, which is used to train a large amount of game level data. The generative pre-training model performs unsupervised learning on large-scale data sets to learn the statistical characteristics and internal laws of the data. After pre-training, the generative pre-training model can be used to generate corresponding results based on the input data. The input data and the generated results are statistically similar to the training data. The sample generation process is autoregressive, and the next part is predicted based on the previous part of the input. This application trains the generative pre-training model based on sample game data to obtain a trained generative pre-training model. Based on the game data, the trained generative pre-training model is called to predict the results of the changes in the game situation, and the optimal strategy for the current situation is obtained, that is, the next operation with the highest probability of the user passing the level.

[0097] The optimal prompt strategy is generated based on the prediction results combined with the current game configuration information. When the user is connected to the Internet, it is optimized through cloud computing. Various data in the game are collected in real time, including player behavior, game progress, and past strategy effect data, and uploaded to the server. The sample game data is continuously updated in the server, and the deep learning model is trained based on the updated sample game data. According to the player's choice and pass rate in the same situation, the priority of each choice in the current situation is verified and optimized to obtain the trained deep learning model. Based on the trained deep learning model, a variety of candidate strategies are generated. Cloud computing simulates the performance of different strategies in the virtual game environment, evaluates their effectiveness and efficiency, and optimizes the candidate strategies based on the evaluation results. The optimal prompt information is selected and fed back to the human-computer interaction interface in real time.

[0098] In the embodiment of the present application, by obtaining the first graphic data of the various first blocks currently displayed in the human-computer interaction interface of the game and the first state information of each first block, the deep learning model is called according to the first graphic data and the first state information to perform operation prediction processing, and the operation prediction result is obtained. The game state and player behavior are analyzed to predict the user's possible optimal operation strategy. Combined with the prompt strategy of deep analysis, the player's satisfaction and stickiness are significantly improved, and the entertainment and attractiveness of the game are enhanced. Prompt information is generated according to the operation prediction results and game configuration information and displayed in the human-computer interaction interface to increase the interactivity of the game. The deep learning model can provide personalized prompts and strategies to improve the user experience based on the player's historical data and operation habits. The generative pre-training model is applied in combination with a deep reinforcement learning strategy to form a self-learning and adaptive prompt engine. By continuously learning player behavior data, the prompt content is dynamically optimized to ensure that the prompt not only adapts to the player's current level, but also can continue to evolve as the player's ability improves. The optimized prompt strategy is fed back to the player in real time in the most appropriate way to ensure the practicality and operability of the prompt information.

[0099] The following is an explanation of an exemplary application of the embodiments of the present application in a practical application scenario.

[0100] In the current game industry, in order to improve user experience and increase user stickiness, game strategy prompt information is usually provided in real time during the game according to the player's current game situation to ensure the player's gaming experience. In the related art, prompts are generated based on predefined fixed rules, lacking the ability to identify and dynamically adjust the personalized needs of players. It is usually impossible to provide real-time personalized prompts based on the real-time status of the game, resulting in poor player experience. The design of the functional framework usually lacks compatibility during expansion, which means that when the game content or mechanism is expanded, the prompt function needs to be redesigned. This limitation increases the difficulty of development. The prompt system usually has a certain delay when responding to player operations, and cannot quickly generate corresponding prompts. This lag may interfere with the player's gaming experience and even affect the overall smoothness of the game.

[0101] The embodiment of the present application trains the level data through a generative pre-training model to obtain a database for strategy prediction. The data in the database is a high-dimensional feature vector processed by the generative pre-training model. The feature vector contains the semantic information of the game data and is used to analyze and process the complex situation of the game. The player data and the current situation information are input into the machine learning model in real time. Through the deep reinforcement learning algorithm, the best strategy that the player should take in each game state is predicted, and prompt information is generated based on the prediction results, and the prompt information is displayed to the user in a timely manner in the most appropriate way.

[0102] The following is a description with reference to the accompanying drawings. Figure 4 , Figure 4 4 is a schematic diagram of a fourth process flow of a method for processing game data provided in an embodiment of the present application. The execution subject may be a terminal device, a server, or a combination of the two. Figure 1 As an example, the server in the example is used as the execution subject. Figure 4 The steps shown are explained in detail.

[0103] In step 401, data is collected through level simulation, and the collected data is preprocessed to obtain a training data set.

[0104] For example, a database of massive data is established to record and analyze the correctness of handling complex situations. Based on the premise that the core framework remains unchanged, technical personnel in this field simulate a large number of game levels through self-testing and automatic operation of game levels by editors, and remove the parts that directly cause failure from all initial data that affect the change in the probability of passing the level during the collection and simulation process. The data obtained from the simulated level includes: graphic information (graphic location, status, graphic type), player operation information (graphic exchange, elimination, movement and other operations) and game panel status information (graphic arrangement status, player score, level completion conditions, current status progress). The collected simulated level data is preprocessed to remove noise and irrelevant information in the level data, and the data information is converted into a format suitable for model processing as a training data set for the model.

[0105] In step 402, a machine learning model is called to make predictions based on the training data set to obtain the optimal strategy.

[0106] For example, the machine learning model is a generative pre-training model, which is used to train a large amount of game level data. The generative pre-training model performs unsupervised learning on large-scale data sets to learn the statistical characteristics and internal laws of the data. After pre-training, the generative pre-training model can predict the next part based on the previous part of the input. Based on the training data set, the generative pre-training model is called to predict the results of the changes in the game sample data to obtain the optimal strategy for the current situation.

[0107] In some embodiments, see Figure 5 , Figure 5 It is a structural diagram of the generative pre-training model provided in an embodiment of the present application; the generative pre-training model 501 is composed of an input layer 5011, an embedding layer 5012, a generative pre-training layer 5013, a reinforcement learning combination layer 5014, a decoding layer 5015 and an output layer 5016.

[0108] For example, the preprocessed training data set is input into the input layer 5011. After receiving the input level data, the input layer 5011 standardizes the data to adapt to the subsequent model layers. The embedding layer 5012 captures the semantic information of the discrete input data through the embedding matrix and converts the discrete data into a continuous vector representation. The generative pre-training layer 5013 has a multi-layer attention mechanism. It extracts features through multi-layer attention concentration, obtains the relationship between different values ​​in the input data, and obtains the level data features. Based on the level data features, the reinforcement learning layer 5014 is called to optimize the strategy. Through the deep learning algorithm, the strategy is adjusted according to the historical operation information and clearance rate of the player in the historical situation, and the priority of different operation choices is evaluated. The processed features are decoded by the decoding layer 5015 to generate the predicted results, and the predicted results of the situation changes are output through the fully connected layer. The output layer 5016 outputs the predicted results of the situation changes and the optimal strategy through the activation function (for example: softmax).

[0109] Continue to see Figure 4 In step 403, the optimal strategy and game data are used to optimize the generative pre-trained model.

[0110] For example, the optimal prompt strategy is generated based on the prediction results, and optimized through cloud computing when the user is connected to the Internet. Various data in the game are collected in real time, including player behavior, game progress, and past strategy effect data, and uploaded to the server. The machine learning model is deployed in the cloud, and the collected data is used for continuous training and updating. According to the player's choices and pass rate in the same situation, the priority of each choice in the current situation is continuously verified and optimized, the model is improved, and the ability to cope with different game situations is enhanced and expanded. Based on the updated model, a variety of candidate strategies are generated. Cloud computing simulates the performance of different strategies in the virtual game environment, evaluates their effectiveness and efficiency, and optimizes the candidate strategies based on the evaluation results. The optimal strategy is selected and fed back to the client in real time to respond to various game situations.

[0111] In step 404, prompt information is generated based on the optimized generative pre-trained model.

[0112] For example, the optimized prompt strategy is fed back to the player in real time in the most appropriate way through the visual interface to ensure the practicality and operability of the prompt. Players can receive prompt information through an intuitive visual interface during the game. These prompt information include but are not limited to presentation in the form of graphics, text or sound to help players quickly understand and take action. Through the prompt information, players can clearly see the current game status analysis, recommended next steps and potential game strategies, thereby improving operational efficiency and gaming experience.

[0113] For example: In the scenario of a graphic elimination game, the basic optimal solution may mean that the system analyzes all the same icons that can be connected on the current board, and recommends that players prioritize connecting those operations that can eliminate multiple icons at a time and get higher scores. Different levels of the game may introduce special rules or mechanisms, which can be: time limits or special icons. The time limit means that players need to complete the game within a limited time. The optimal strategy will take time factors into account when prompting, and recommend players to take strategies with fewer steps but may score slightly lower. Special icons are special icons that can eliminate an entire row or column. The optimal strategy will consider how to effectively use these special icons to get higher scores. It may also include new obstacles or props with special functions.

[0114] In some embodiments, see Figure 6 , Figure 6 is a schematic diagram of a human-computer interaction interface provided in an embodiment of the present application, Figure 6 601, which displays time information 6011 corresponding to the time limit of the level. If the current player does not trigger an operation for a long time, the game determines that the player needs the help of prompt information, and triggers a dynamic prompt icon 602 to remind the user to click the icon to activate the function. After the player triggers the dynamic prompt icon 602, prompt information 603 will be displayed in the graphical interface to quickly give feedback on the user's operation, showing the calculation results for simple or complex situations, that is, the optimal strategy that can be executed in the next step.

[0115] The method for processing game data provided by the embodiment of the present application has the following beneficial effects:

[0116] By applying generative pre-training models combined with deep reinforcement learning strategies, we can dynamically optimize the prompt content by continuously learning player behavior data, ensuring that the prompts not only adapt to the player's current level, but also evolve as the player's ability improves. Cloud computing technology is used to transfer complex computing and data processing tasks to the cloud, reduce the burden on local devices, improve the response speed and accuracy of the prompt system, quickly generate efficient prompts, and ensure that players' needs can be met in the shortest time. Local computing and cloud computing resources are dynamically allocated according to the amount of computing, ensuring a smooth user experience and optimizing resource utilization efficiency under different device and network conditions. Through real-time and efficient prompt information, players can better master game skills and explore more game strategies.

[0117] The following is a description of an exemplary structure of the game data processing device 455 provided in the embodiment of the present application implemented as a software module. In some embodiments, Figure 2As shown, the software modules in the processing device 455 of the game data stored in the memory 450 may include: a data acquisition module 4551, which is used to obtain the first graphic data of multiple first tiles currently displayed in the human-computer interaction interface of the game and the first state information of each first tile; wherein, the number of the same type of first tiles is multiple, and the graphics corresponding to the same type of first tiles are the same, and the state information includes: a to-be-selected state and a prohibited selection state; an operation prediction module 4552, which is used to call the deep learning model to perform operation prediction processing according to the first graphic data and the first state information to obtain an operation prediction result; generate prompt information according to the operation prediction result and the game configuration information; display the prompt information in a preconfigured area in the human-computer interaction interface, wherein the prompt information is used to characterize the position of a preset number of second tiles, and the second tiles are the same type of first tiles in the to-be-selected state for performing a selection operation, and when the preset number of second tiles are selected, the preset number of second tiles are hidden.

[0118] In some embodiments, before obtaining the first graphic data of multiple first tiles currently displayed in the human-computer interaction interface of the game and the first status information of each first tile, the data acquisition module 4551 is also used to obtain historical game data, wherein the historical game data includes: the first graphic data of the first tile in each game, the selection operation for each first tile and the game result of each game; denoising the historical game data to obtain sample game data; training the deep learning model based on the sample game data to obtain a trained deep learning model.

[0119] In some embodiments, the operation prediction module 4552 is further used to call the deep learning model to perform the following processing according to the first graphic data and the first state information: perform feature extraction processing on the first graphic data and the first state information to obtain current game features; perform operation strategy prediction processing based on the current game features to obtain operation prediction results. In some embodiments, the operation prediction module 4552 is further used to perform conversion processing on the first graphic data and the first state information to obtain a first embedded vector representation, wherein the first embedded vector representation is represented as a sequence of multiple feature vectors; perform normalization processing on the first embedded vector representation to obtain a second embedded vector representation; perform multiple feature extraction processing on the second embedded vector representation to obtain current game features, wherein the input of each feature extraction processing is the output of the previous feature extraction processing.

[0120] In some embodiments, the operation prediction module 4552 is also used to perform prediction processing based on the current game characteristics to obtain different types of third tiles in the first tile to be selected; for each third tile, multiple prediction processing is performed based on the current game characteristics and the same third tiles until a preset number of the same third tiles are obtained, wherein the input of each prediction processing includes: the same third tiles that have been predicted, and the current game characteristics updated based on the predicted third tiles of the same type; one is selected from each preset number of third tiles as the second tile; the prediction order of the second tiles and the second graphic data of each second tile are used as the operation prediction result, wherein the prediction order is used as the selection order of the second tiles.

[0121] In some embodiments, the operation prediction module 4552 is also used to predict the game success rate of a preset number of the same third tiles for each third tile based on historical game operation results, wherein the game winning rate represents the success probability of the current game after the preset number of the same third tiles are hidden; and the preset number of the same third tiles with the highest game success probability are selected as the second tile.

[0122] In some embodiments, the operation prediction results include: the selection order of the second tiles and the second graphic data of each second tile; the operation prediction module 4552 is also used to use the pre-configured icon as the prompt icon for each second tile; according to the selection order, text generation processing is performed based on the second graphic data of each second tile and the game configuration information to obtain the prompt text corresponding to each second tile; according to the selection order, the prompt text of each second tile and the display order of the prompt icon are marked; and the marked prompt text and prompt icon are used as prompt information.

[0123] The embodiment of the present application provides a computer program product, which includes a computer program or a computer executable instruction, and the computer program or the computer executable instruction is stored in a computer-readable storage medium. The processor of the electronic device reads the computer executable instruction from the computer-readable storage medium, and the processor executes the computer executable instruction, so that the electronic device executes the above-mentioned method for processing game data in the embodiment of the present application.

[0124] The present application embodiment provides a computer-readable storage medium in which computer executable instructions or computer programs are stored. When the computer executable instructions or computer programs are executed by a processor, the processor will execute the game data processing method provided by the present application embodiment, for example, Figure 3A The game data processing method is shown.

[0125] In some embodiments, the computer-readable storage medium may be a memory such as RAM, ROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or may be various devices including one or any combination of the above memories.

[0126] In some embodiments, computer executable instructions may be in the form of a program, software, software module, script or code, written in any form of programming language (including compiled or interpreted languages, or declarative or procedural languages), and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine or other unit suitable for use in a computing environment.

[0127] As an example, computer-executable instructions may, but need not, correspond to a file in a file system, may be stored as part of a file that stores other programs or data, such as in one or more scripts in a HyperText Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files storing one or more modules, subroutines, or code portions).

[0128] As an example, computer executable instructions may be deployed to be executed on one electronic device, or on multiple electronic devices located at one site, or on multiple electronic devices distributed at multiple sites and interconnected by a communication network.

[0129] To summarize, through the embodiments of the present application, a deep learning model is called to perform operation prediction on the graphic data and status information displayed in the human-computer interaction interface, and the user's possible optimal operation strategy is predicted to obtain the operation prediction result. According to the operation prediction result and the game configuration information, prompt information is generated in real time and displayed to the player, thereby improving the user experience and ensuring the practicality and operability of the prompt information.

[0130] The above is only an embodiment of the present application and is not intended to limit the protection scope of the present application. Any modifications, equivalent substitutions and improvements made within the spirit and scope of the present application are included in the protection scope of the present application.

Claims

1. A method for processing game data, characterized in that: The method comprises: Acquire first graphic data of multiple first blocks currently displayed in the human-computer interaction interface of the game and first state information of each of the first blocks; There are multiple first blocks of the same type, and the graphics corresponding to the first blocks of the same type are the same, and the state information includes: a waiting state and a prohibited selection state; Calling a deep learning model to perform operation prediction processing according to the first graphic data and the first state information to obtain an operation prediction result; Generate prompt information according to the operation prediction result and game configuration information; The prompt information is displayed in a preconfigured area in the human-computer interaction interface, wherein the prompt information is used to represent the position of a preset number of second tiles, the second tiles are the same type of first tiles in the to-be-selected state for performing a selection operation, and when the preset number of second tiles are selected, the preset number of second tiles are hidden.

2. The method according to claim 1, characterized in that The calling of the deep learning model to perform operation prediction processing according to the first graphic data and the first state information to obtain the operation prediction result includes: According to the first graphic data and the first state information, the deep learning model is called to perform the following processing: Performing feature extraction processing on the first graphic data and the first state information to obtain current game features; An operation strategy prediction process is performed based on the current game features to obtain an operation prediction result.

3. The method according to claim 2, characterized in that The performing feature extraction processing on the first graphic data and the first state information to obtain current game features includes: Converting the first graphic data and the first state information to obtain a first embedded vector representation, wherein the first embedded vector representation is represented as a sequence consisting of a plurality of feature vectors; Normalizing the first embedded vector representation to obtain a second embedded vector representation; The second embedding vector representation is represented by performing feature extraction processing multiple times to obtain the current game features, wherein the input of each feature extraction processing is the output of the previous feature extraction processing.

4. The method according to claim 2, characterized in that: The performing operation strategy prediction processing based on the current game features to obtain an operation prediction result includes: Performing prediction processing based on the current game feature to obtain a third block of a different type in the first block in the to-be-selected state; For each of the third tiles, multiple prediction processes are performed based on the current game features and the third tiles of the same type, until a preset number of third tiles of the same type are obtained, wherein the input of each prediction process includes: the third tiles of the same type that have been predicted, and the current game features that are updated based on the third tiles of the same type that have been predicted; Selecting one of the preset number of third blocks as the second block; The predicted order of the second tiles and the second graphic data of each of the second tiles are used as the operation prediction result, wherein the predicted order is used as the selection order of the second tiles.

5. The method according to claim 4, characterized in that The selecting one from each of the preset number of third blocks as the second block comprises: For each type of the third tiles, predict the game success rate of the preset number of the same type of third tiles according to historical game operation results, wherein the game win rate represents the success probability of the current game after the preset number of the same type of third tiles are hidden; A preset number of third tiles of the same type with the highest probability of success of the game are selected as the second tiles.

6. The method according to claim 1, characterized in that The operation prediction result includes: a selection order of the second blocks and second graphic data of each of the second blocks; The generating prompt information according to the operation prediction result and the game configuration information includes: Using a preconfigured icon as a prompt icon for each of the second tiles; According to the selection order, text generation processing is performed based on the second graphic data of each second block and the game configuration information to obtain a prompt text corresponding to each second block; According to the selection order, marking the prompt text and the display order of the prompt icon of each second block; The marked prompt text and the prompt icon are used as the prompt information.

7. The method according to any one of claims 1 to 6, characterized in that: Before obtaining the first graphic data of the multiple first blocks currently displayed in the human-computer interaction interface of the game and the first state information of each of the first blocks, the method further includes: Acquire historical game data, wherein the historical game data includes: first graphic data of the first block in each game, a selection operation for each first block, and a game result of each game; Performing denoising processing on the historical game data to obtain sample game data; The deep learning model is trained based on the sample game data to obtain the trained deep learning model.

8. A device for processing game data, characterized in that: The device comprises: A data acquisition module, used to acquire first graphic data of multiple first blocks currently displayed in the human-computer interaction interface of the game and first state information of each first block; wherein the number of the same first blocks is multiple, the graphics corresponding to the same first blocks are the same, and the state information includes: a waiting state and a prohibited selection state; An operation prediction module is used to call a deep learning model to perform operation prediction processing according to the first graphic data and the first state information to obtain an operation prediction result; generate prompt information according to the operation prediction result and game configuration information; and display the prompt information in a preconfigured area in the human-computer interaction interface, wherein the prompt information is used to represent the position of a preset number of second tiles, the second tiles are the same type of first tiles in the to-be-selected state for performing a selection operation, and when the preset number of second tiles are selected, the preset number of second tiles are hidden.

9. An electronic device, characterized in that: The electronic device comprises: A memory for storing computer executable instructions or computer programs; A processor is used to implement the method for processing game data according to any one of claims 1 to 7 when executing the computer executable instructions or computer programs stored in the memory.

10. A computer-readable storage medium storing computer-executable instructions or a computer program, characterized in that: When the computer executable instructions or computer program are executed by a processor, the method for processing game data according to any one of claims 1 to 7 is implemented.

Images