A game prop operation method and device and a computer readable storage medium

Abstract

The application discloses a game prop operation method and device and a computer readable storage medium, which collects game data, behavior data and environment data of a game player; receives a game prop purchase request, generates a game prop purchase suggestion according to the game data and the behavior data of the game player; receives a game prop use request, customizes and displays a use strategy of the game prop according to the behavior data of the game player; receives a game prop effect generation request, and generates an adapted effect according to the environment data of a game process in which the game player is located; from the purchase of the game prop to the use and then to the presentation of the effect, personalized recommendation, customization and flexible presentation can be realized according to the specific needs and preferences of the game player, and the flexibility and interest of the use of the game prop can be improved.

Description

Technical Field

[0001] This invention relates to the field of game design, and more particularly to a method, apparatus, and computer-readable storage medium for operating game props. Background Technology

[0002] In many existing games, purchasing items typically requires players to locate the corresponding item in the game's shop interface, click the purchase button, and complete the transaction by confirming the payment method (such as using in-game currency or making a purchase with real money). After a successful purchase, the item appears in the player's inventory. The player then needs to locate the item in the inventory interface, select the "use" option, and then choose the target (if the item has a specific target). However, this process is often cumbersome, involving switching between multiple interfaces and operational steps, which can easily confuse and tire players.

[0003] For example, some games have complex item shop categorizations, which may require players to spend a lot of time searching for specific items; the item arrangement in the inventory is fixed, making it difficult to quickly locate and operate items. Moreover, the usage restrictions and effects of items are not clearly explained, which may lead players to use items incorrectly without understanding the detailed rules, thus affecting the game experience.

[0004] In the process of using items, most games simply restrict the use of items based on the character's level, such as low-level characters being unable to use items with high-level requirements. However, they give little consideration to the impact of different character states (such as combat state, mission state, etc.) on the effects of item use, lacking in-depth logical connections.

[0005] For example, in some games, characters can freely use healing items while in combat, without considering combat balance and strategy; for some special items, their effects remain unchanged regardless of the character's current mission objective or the scene they are in, and cannot be dynamically adjusted according to the game situation, making the gameplay relatively monotonous. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method, device and computer-readable storage medium for operating game props, which can improve the flexibility and fun of using game props.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: A method for operating a game item, including the following steps: Collect game data, behavioral data, and environmental data from game players during gameplay; Receive game item purchase requests and generate game item purchase suggestions based on the game player's game data and behavior data; Receive requests to use game items, and customize and display strategies for using the game items based on the player's behavior data. Receive a request to generate game item effects, and generate appropriate effects based on the environmental data of the game player's current game process.

[0008] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: An operating device for a game prop includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the above-described operating method for the game prop.

[0009] To solve the above-mentioned technical problems, another technical solution adopted by the present invention is as follows: A computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, implement the steps of the above-described method for operating a game item.

[0010] The beneficial effects of this invention are as follows: it collects game data, behavioral data, and environmental data of game players; when game players purchase game items, it generates game item purchase suggestions based on the game player's game data and behavioral data; when game players use game items, it customizes and displays the usage strategy of the game items based on the game player's behavioral data; and during the game, it generates appropriate game item effects based on the environmental data of the game player's current game process. From the purchase of game items to their use and the presentation of their effects, it can achieve personalized recommendations, customization, and flexible presentation according to the specific needs and preferences of game players, thereby improving the flexibility and fun of using game items. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating the steps of an operation method for a game prop according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the operating device for a game prop according to an embodiment of the present invention. Detailed Implementation

[0012] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0013] Please refer to Figure 1 A method for operating a game item, including the following steps: Collect game data, behavioral data, and environmental data from game players during gameplay; Receive game item purchase requests and generate game item purchase suggestions based on the game player's game data and behavior data; Receive requests to use game items, and customize and display strategies for using the game items based on the player's behavior data. Receive a request to generate game item effects, and generate appropriate effects based on the environmental data of the game player's current game process.

[0014] As described above, the beneficial effects of this invention are as follows: it collects game data, behavioral data, and environmental data of game players; when game players purchase game items, it generates game item purchase suggestions based on the game players' game data and behavioral data; when game players use game items, it customizes and displays the usage strategy of the game items based on the game players' behavioral data; and during the game, it generates appropriate game item effects based on the environmental data of the game players' current game process. From the purchase of game items to their use and the presentation of their effects, it can achieve personalized recommendations, customization, and flexible presentation according to the specific needs and preferences of game players, realizing deep interaction between game characters and game items, and improving the flexibility and fun of using game items.

[0015] Furthermore, it also includes the following steps: Receive social group operation requests; Perform the corresponding operation based on the social group operation request; The game item growth value for each player is determined based on the operation results.

[0016] As described above, by setting up social groups and performing various operations on these groups, the game items corresponding to the players can be increased based on the operational structure, further enhancing the interaction between players and game items and improving the flexibility of game item usage.

[0017] Furthermore, a behavioral analysis model is constructed using a machine learning framework; Features and corresponding labels are defined according to different analysis objectives, and features are extracted periodically from the game data, behavior data and environmental data of the game players collected. The behavior analysis model is trained and updated based on the extracted features and corresponding labels; The corresponding behavioral analysis model is determined based on the type of request received regarding game items; The corresponding behavior analysis model is used to execute corresponding game item operations, which include game item purchase suggestions, game item usage strategies, and game item effects.

[0018] As described above, a behavior analysis model is constructed using a machine learning framework. Features and corresponding labels are defined according to different analysis objectives. The corresponding features are extracted from the collected data for training. The behavior analysis model is determined according to the type of game item request received to execute the corresponding game item operation, thus ensuring the efficiency and accuracy of the executed game item operation.

[0019] Furthermore, the game data includes the player's character's growth progress, game level, and pet training status, and the behavioral data includes the player's item usage history; The process of generating game item purchase suggestions based on the game player's game data and behavior data includes: The correlation between the items used and the corresponding game character's growth progress, game level, or pet training status is determined based on the player's item usage history. Game item purchase suggestions are generated based on the correlation.

[0020] As described above, by determining the correlation between the items used and the corresponding game character's growth progress, game level, or pet training status based on the game player's item usage history, and generating game item purchase suggestions based on the correlation, the accuracy of the generated game item purchase suggestions is ensured, thus meeting the actual needs of game players.

[0021] Furthermore, the behavioral data includes the game player's character development goals; The strategy for customizing and displaying the use of game items based on the game player's behavior data includes: Based on the player's character development goals, determine the game items that match those goals and assign priority or proportion to these items.

[0022] As described above, by determining game items that match the game player's character development goals and prioritizing or allocating these items, personalized customization of game item usage is achieved, satisfying the individual needs of each player.

[0023] Furthermore, the environmental data includes weather data, map area data, and time data; The adaptation effect generated based on the environmental data of the game player's game process includes: The game level upgrade effect is generated based on the map area data of the game player's current game process; The game item bonus effect is generated based on the weather data during the game. The game item usage rules and effects are generated based on the time data of the player's game process.

[0024] As described above, the ability to generate corresponding game item effects based on different environmental data in the game scene further enhances the game's fun.

[0025] Furthermore, the behavioral data also includes social activity data; The social activity data includes social activity level, social relationship intimacy, and social group goals; The recommendation weight of different game items is dynamically adjusted based on the social activity level. A model of social relationship intimacy among game players is constructed based on the aforementioned social relationship intimacy. The recommendation priority of game items is determined based on the aforementioned social relationship intimacy model; Game items are recommended based on the alignment between the player's game data and the social goals of their social group.

[0026] As described above, by leveraging social activity data, game items can be recommended based on a player's social activity level, their close relationships with other social members, and the goals of their social group. This can increase players' enthusiasm for social interaction and further enhance the game's enjoyment.

[0027] Furthermore, when generating effects that adapt to game items: Define a state space and an action space. The state space covers various environmental data of the game process, and the action space includes various possible attribute change strategies of game items. A deep Q-network is used to reinforce the machine learning framework. Through training, the constructed behavioral learning model learns to dynamically formulate the optimal strategy for changing the attributes of game items based on the game environment and the state of the player.

[0028] As described above, defining a definition space that encompasses various environmental data during the game process and an action space that includes various possible attribute change strategies for game items, and employing a deep Q-network to reinforce the machine learning framework, enables the constructed behavioral model to quickly and accurately formulate the optimal attribute change strategy for game items based on the game environment and the player's state.

[0029] Please refer to Figure 2 An operating device for a game prop includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the above-mentioned operating method for a game prop.

[0030] A computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, implement the steps of the above-described method for operating a game item.

[0031] The method, apparatus, and computer-readable storage medium for operating game props described in this application are applicable to prop operations in various game processes. The following description illustrates the implementation methods: In one alternative implementation, such as Figure 1 As shown, a method for operating a game item includes the following steps: S1. Collect game data, behavioral data, and environmental data during the game process from game players; S2. Receive a game item purchase request and generate a game item purchase suggestion based on the game player's game data and behavior data; S3. Receive a request to use game items, and customize and display the strategy for using the game items based on the game player's behavior data; S4. Receive a request to generate game item effects, and generate appropriate effects based on the environmental data of the game player's current game process.

[0032] The game data includes the player's character's growth progress, game level, and pet training status, while the behavioral data includes the player's item usage history. The process of generating game item purchase suggestions based on the game player's game data and behavior data includes: The correlation between the items used and the corresponding game character's growth progress, game level, or pet training status is determined based on the player's item usage history. Game item purchase suggestions are generated based on the correlation.

[0033] For example, the Magic Stone Shop interface incorporates an intelligent recommendation algorithm based on player game data and behavior analysis. The system monitors multi-dimensional data in real time, including the player character's growth progress, current Aura level, pet training status, and past item usage habits. For instance, if a player frequently uses a specific skill in battle, and that skill has the potential to synergize with Aura level increases, the system will prioritize recommending Experience Orb packs that significantly boost related Aura attributes. Simultaneously, it considers external factors such as in-game event cycles and the overall strength of the player's server to provide purchase suggestions best suited to the current game situation.

[0034] When players open the Magic Stone Shop, the recommendation interface prominently and dynamically displays recommended Energy Core Packs, along with detailed reasons for the recommendation and an analysis of expected effects, such as "Based on your character's skill usage frequency, this pack will significantly enhance your special ability performance in [specific skill area], and is expected to increase the skill damage by [X]%." Players can use these intelligent recommendations to more accurately choose packs suitable for their character's development, avoiding blind purchases.

[0035] In practice: 1. Data Collection and Preprocessing (Data Acquisition, Cleaning, Transformation - ETL Process) The server collects player game data through various methods, including but not limited to: retrieving basic character information from the database (such as level, class, power level, pet list and level, etc.), combat data (such as skill usage frequency, damage output, damage taken, combat win rate, etc.), and item usage history (records of past purchases and uses of energy cores, including usage time, user, and effects). The collected data is cleaned to remove invalid or erroneous data (such as abnormal combat data records) and converted into a format suitable for analysis (such as converting raw database records into star or snowflake schemas in a data warehouse). This process involves ETL (Extract, Transform, Load) operations, using ETL tools (such as Apache Sqoop for data extraction and Apache Hive for data transformation and storage) to efficiently process the data.

[0036] At the same time, the system monitors the status of in-game activities in real time (such as the type of activity currently in progress, the duration of the activity, the number of players participating in the activity, etc.) and overall server data (such as server load, number of online players, distribution of players of different classes, etc.). This external environmental data is integrated with the player's personal data to provide a comprehensive data foundation for intelligent recommendations.

[0037] 2. Recommendation Algorithm Calculation (Application of Machine Learning Algorithms - Collaborative Filtering Algorithm and Matrix Factorization Technique) Machine learning algorithms (such as collaborative filtering, content-based recommendation, or hybrid recommendation algorithms) are used to analyze and compute the preprocessed data. Collaborative filtering analyzes the behavioral patterns of player groups to find other players with similar gameplay behaviors to the target player. Based on these similar players' purchase and usage of Energy Core Packs, it recommends packs that the target player might be interested in. In the implementation of collaborative filtering, matrix factorization techniques (such as Singular Value Decomposition (SVD)) are often used to handle large-scale user-item matrices, reducing data dimensionality and improving computational efficiency, thus finding similar users and recommended packs more quickly and accurately. Content-based recommendation algorithms match the attributes of Energy Core Packs (such as the type of Qi attribute boosted, applicable character class, and correlation with specific skills) with the needs and characteristics of the target player's character (such as the character's current Qi weakness and commonly used skill characteristics) to select suitable recommended packs.

[0038] The hybrid recommendation algorithm combines the results of collaborative filtering and content-based recommendation to calculate a recommendation score for each Energy Core pack. For example, a content-based score is calculated based on the player's skill usage frequency and the energy core's enhancement effect on the skill's related special attributes. Simultaneously, a collaborative filtering score is calculated by combining the purchase frequency of similar players for that pack. The two scores are then weighted and summed to obtain the final recommendation score. The packs are then ranked according to their recommendation scores, and the top-ranked packs are selected as the recommended results.

[0039] 3. Recommendation results push and display (network communication - WebSocket push, interface update - UI refresh mechanism) The server encapsulates the recommendation results (including gift pack information, reasons for recommendation, expected effects, etc.) into data in a specific format (such as JSON format) and pushes the data to the client via network communication protocols (such as WebSocket or HTTP push technology, taking WebSocket as an example here). After receiving the recommendation data, the client updates and displays it in real time on the Magic Stone Shop interface. Recommended gift packs are displayed in a prominent manner (such as using special color markings, increasing icon size, or adding dynamic effects), and detailed reasons for recommendation are displayed next to them (such as "Based on your recent high frequency of use of [skill name] in battle, this gift pack can significantly improve the related Qi Power attribute of this skill, and is expected to increase skill damage by [X]%") and a preview of the expected effects (such as showing the attribute changes after the Qi Power level is increased in the form of animation or charts). This involves the client's UI refresh mechanism, which updates the interface display by calling the refresh function of the graphical user interface framework (such as Unity's UGUI or Android's XML layout) to guide players in making purchase decisions. The behavioral data includes the game player's character development goals; The customizing and displaying the usage strategy of the game prop according to the behavior data of the game player includes: Determine game props adapted to the game character development target according to the game character development target of the game player, and perform priority setting or proportion allocation on the adapted game props.

[0040] For example, when operating to use an experience value orb in the backpack, a personalized usage plan customization function is provided for the player. Players can set the usage priority and allocation strategy of energy cores according to their own character development goals (such as focusing on different directions including PVP combat, PVE dungeon strategy, phantom beast breeding, etc.). For example, for players focusing on PVP, the priority can be set to increase the mystery power levels related to key attributes of the character in combat (such as critical hit rate, dodge rate, etc.); for players focusing on phantom beast breeding, a usage plan for allocating the experience value orb to different phantom beasts at a certain proportion can be customized.

[0041] The system provides intuitive operation guidance and effect preview on the usage interface according to the personalized plan set by the player. When the player drags the experience value orb to the usage box, it will display the detailed mystery power level improvement values, attribute changes after use according to the customized plan, and the performance prediction of the character or phantom beast in a specific game scenario (such as the selected PVP mode or PVE dungeon type), helping the player better plan and adjust the character development path.

[0042] In specific implementation: 1. Player target setting and data storage (user input - event listening and callback, database operation - transaction processing) The client provides a setting entry for the player to customize the personalized usage plan in the backpack or character management interface. Players can set their own character development goals (such as selecting options like "PVP combat enhancement", "PVE dungeon breakthrough", "phantom beast limit breeding", or entering a custom description of the development direction) and specific usage priority and allocation strategies through components such as drop-down menus, radio buttons or text input boxes of the graphical user interface (GUI). For example, in the direction of phantom beast breeding, set the proportion of energy cores allocated to different phantom beasts, such as Phantom Beast A: 40%, Phantom Beast B: 30%, Phantom Beast C: 30%. The client listens to the player's input operation through an event listening mechanism (such as the addEventListener function of JavaScript or the event delegation of C), and when the player completes the setting and submits, a callback function is triggered to collect and encapsulate the data set by the player, and then the data is sent to the server through network communication.

[0043] After receiving the request, the server stores the player's personalized settings data in a dedicated table in the database (such as a player personalization settings table), associating it with the player's character information to ensure persistent data storage. During database operations, transaction processing (such as MySQL's transaction mechanism) is used to guarantee data consistency and integrity. This means that either all data insertion or update operations complete successfully, or all are rolled back, to prevent situations where some data updates succeed while others fail. This ensures that these personalized settings can be retrieved and applied later when using experience orbs.

[0044] 2. Use dynamic interface adjustments and effect previews (interface rendering - graphics pipeline, computational logic - algorithm optimization) When a player selects to use an Energy Core in their inventory, the client retrieves the player's personalized settings data from the server and dynamically adjusts the display and operation logic of the user interface based on this data. Regarding interface rendering, operations involving the graphics pipeline (such as OpenGL or DirectX) are performed. Based on different priorities and allocation strategies, the drawing order, color, transparency, and other attributes of graphical elements are adjusted to highlight relevant information. For example, based on the player's set usage priority, the usage box prioritizes displaying character or pet options that match the priority level, using different colors or icons to indicate their priority. Simultaneously, according to the allocation strategy, suggested values ​​or proportions of Energy Cores allocated to different objects are displayed on the interface, allowing players to fine-tune them as needed.

[0045] The client calculates and previews the effects of the energy core in real time, based on the currently selected user and its attributes, combined with the server-side calculation interface or a locally predefined calculation model. The calculation process involves determining the base enhancement value based on the energy core's level and type, then adjusting it according to the player's personalized strategy (e.g., weighted calculation of the power level increase related to key combat attributes for PVP combat enhancement goals). This results in the final power level increase value, attribute changes (e.g., specific increases in attack power, defense power, critical hit rate, etc.), and performance predictions in specific game scenarios (e.g., the selected PVP mode or PVE dungeon type) (e.g., the expected percentage increase in win rate in PVP, the expected reduction in clear time in a specific dungeon). To improve calculation efficiency, algorithm optimization techniques (e.g., dynamic programming algorithms to optimize attribute bonus calculations and avoid redundant calculations) can be used in the calculation logic. These preview effects are displayed intuitively on the user interface (e.g., displayed as numbers, charts, or text descriptions in specific areas of the interface) to help players make more informed usage decisions.

[0046] The environmental data includes weather data, map area data, and time data; The adaptation effect generated based on the environmental data of the game player's game process includes: The game level upgrade effect is generated based on the map area data of the game player's current game process; The game item bonus effect is generated based on the weather data during the game. The game item usage rules and effects are generated based on the time data of the player's game process.

[0047] For example, the game world's environmental factors (such as weather changes, map area characteristics, and time cycles) are deeply linked to the effects of using energy cores. For instance, using an energy core in specific mysterious map areas (such as ancient Qili Ruins) will increase the Qili level by an additional percentage, and may also trigger special temporary attribute bonuses (such as granting the character additional elemental resistance when used in the ruins). Different weather conditions will also affect the effects; for example, using it during thunderstorms has a chance to grant the summoned beast a lightning-attribute attack bonus for several battles.

[0048] In terms of time cycles, the rules and effects of using Energy Cores will change depending on the game's day-night cycle or special festival events. For example, using Experience Orbs on "Power Awakening Day" (a specific festival) will grant double the Power level boost and offer a chance to unlock hidden character appearances or skill effects. The introduction of these dynamic environmental factors requires players to choose the best time to use items based on the real-time state of the game world, increasing the game's strategy and fun.

[0049] In practice: 1. Environmental factor monitoring and data acquisition (sensor technology - virtual sensor simulation, game engine interface - callback functions and event-driven) The game server interacts with the game client in real time through the game engine's interface to obtain environmental factor data from the game world. For weather changes, the game client uses a built-in weather simulation system and sensor technology (such as virtual sensors simulating raindrops and light, simulating the generation and changes of sensor data through code) to monitor weather conditions (such as sunny, rainy, thunderstorms, etc.) and transmits the weather data to the server in real time. During this process, the game engine uses callback functions and event-driven mechanisms (such as Unity's event system) to trigger corresponding callback functions when the weather condition changes, sending the weather data to the server. Map area characteristic data is determined by the game engine based on the player's location on the map, transmitting the map area type (such as normal area, Qili Ruins area, Mysterious Forest area, etc.) and related attributes (such as the area's elemental energy intensity, terrain features, etc.) to the server.

[0050] The time-cycle data is obtained from the server's system clock. Simultaneously, the server integrates with the game's event scheduling system to determine the current time within a specific in-game period (e.g., a regular day, a holiday event, a specific dungeon opening period). The server consolidates and stores this environmental data for effect calculations when players use energy cores.

[0051] 2. Effect Calculation and Application (Algorithm Model - Weighted Calculation and Random Number Generation, Database Update - Optimistic Locking Mechanism) When a player uses an Energy Core, the server calculates its actual effect based on current environmental data and the attributes of the Experience Orb, following a predefined algorithm model. For example, when using an Experience Orb in the Qili Ruins area, the algorithm calculates an additional Qili level increase percentage (e.g., an additional 20% increase) based on the ruins' attributes (such as a Mystic Energy bonus) and the Experience Orb's base increase value. It also determines potential special temporary attribute bonuses (e.g., adding elemental resistance to the character based on the ruins' elemental attributes, with the resistance value calculated based on both the ruins' difficulty level and the Experience Orb's level). The calculation process utilizes weighted calculations (e.g., assigning different weights based on the importance of environmental factors to calculate the final effect) and random number generation (e.g., using Java's Random class to randomly adjust attribute bonus values ​​within a certain range, increasing the game's unpredictability).

[0052] After the calculation is complete, the server updates the attribute data of the player character or pet, raising the Qi Power level to the new calculated value and adding corresponding temporary attribute bonuses. During the database update process, to avoid concurrent update conflicts, an optimistic locking mechanism is used (e.g., adding a version number field to the database table, checking if the version number is consistent during the update, and if inconsistent, rolling back and re-acquiring the latest data for calculation and update). At the same time, the usage record (including usage time, location, environmental factors, Qi Power level and attribute changes before and after usage, etc.) is stored in the database for subsequent data analysis and player queries.

[0053] 3. Client feedback and display (network communication - HTTP requests / responses, graphics rendering - animation system and particle effects) The server sends the updated character or pet attribute data and usage effect information (such as the power level increase, temporary attribute bonus type and value, and description of triggered special effects) to the client via network communication protocols (such as HTTP request / response, encapsulating the data in the HTTP response body and returning it to the client). Upon receiving the data, the client displays real-time feedback on the game interface. For example, the character or pet's power level is updated to the new value, and a temporary attribute bonus icon and value (such as "Lightning Resistance +20") are displayed on the character or pet's status panel. The triggered special effects are also displayed with animations or special effects (such as playing a character level-up animation through the game engine's animation system, or using a particle effect system to generate a mysterious flashing light effect to indicate the acquisition of a special bonus), allowing players to intuitively experience the impact of environmental factors on the energy core's usage effect.

[0054] In another alternative implementation, the following steps are also included: Receive social group operation requests; Perform the corresponding operation based on the social group operation request; The game item growth value for each player is determined based on the operation results.

[0055] Specifically, a mechanism for acquiring and sharing Energy Cores based on social interaction will be established. Players can join or create in-game social groups (such as guilds, teams, etc.) and earn additional Energy Core rewards by participating in group activities (such as guild dungeons, team battles, etc.). Within social groups, members can gift Energy Cores to each other, promoting social interaction and cooperation among players.

[0056] A social leaderboard is introduced, ranking players based on their contributions to social groups (such as helping other members improve their character's strength, and their active participation in group activities). Top-ranked players can receive special Energy Core packs or privileges (such as extra daily usage attempts or higher usage effect bonuses). Furthermore, players can invite friends to assist when using Energy Cores; the more friends who assist, the more significant the effect within a certain range (e.g., friend assistance can provide a small additional boost to Qi Power level or attribute bonuses), encouraging players to actively socialize and grow together.

[0057] In practice: 1. Social group activities and reward calculation (activity logic - state machine design, reward system - probability distribution model) The game server maintains the data structures and logic related to social groups (such as guilds and teams). When a social group organizes or participates in a specific activity (such as guild dungeons or team battles), the server calculates rewards based on the activity rules and the performance of participating members. During the activity, the server uses a state machine design pattern (e.g., using a finite state machine to manage different phases of the activity, such as the preparation phase, the activity phase, and the end phase, each with different behaviors and rules) to control the activity logic and record each member's contribution data (e.g., damage dealt, damage taken, healing received, and number of team collaboration actions in guild dungeons). After the activity ends, the server calculates the amount of energy cores each member should receive based on this contribution data. The reward calculation may involve a probability distribution model (e.g., using a normal distribution model to allocate different probabilities of receiving different amounts of rewards based on the member's contribution ranking, with high-performing members having a higher probability of receiving more rewards).

[0058] For social groups that perform exceptionally well (such as achieving high rankings in team competitions or clearing guild dungeons with outstanding results), the server will also issue additional collective rewards to group members (such as extra energy core packs, rare items, etc.). The amount of rewards is calculated and determined based on the evaluation indicators of group performance (such as competition rankings, dungeon clear times, teamwork scores, etc.).

[0059] 2. Item gifting and leaderboard logic (database operations - index optimization, algorithm sorting - heap sort algorithm) When a player chooses to gift Energy Cores to other members in a social group, the client sends a gift request to the server. This request includes the character identifiers of both the gifter and recipient, as well as the number of Experience Orbs to be gifted. Upon receiving the request, the server first verifies whether the gifter has a sufficient number of Experience Orbs. If the verification is successful, the server performs the corresponding operation in the database: reducing the gifter's number of Experience Orbs and increasing the recipient's number, and recording the gift transaction (including the gifting time, character information of both parties, and the number of Experience Orbs gifted). To improve query and update efficiency in database operations, indexes are added to frequently queried fields in relevant database tables (such as the player's inventory table and gifting record table). For example, B-tree indexes are used to optimize the operation of querying the number of Experience Orbs in a player's inventory based on their character identifier.

[0060] The server regularly updates social leaderboard data, calculating rankings based on players' contributions to social groups. Contribution calculation considers multiple factors, such as an individual's performance score in group activities (calculated based on the aforementioned contribution data), activity levels in gifting and receiving energy cores (e.g., number of gifts, number of receipts, total gift amount), and helpful behaviors towards other group members (e.g., quantitative evaluation of assisting other members in completing tasks, providing game guides, etc.). The leaderboard sorting algorithm uses a heap sort algorithm (utilizing the characteristics of a heap data structure to quickly find the maximum or minimum element, constructing a max-heap based on contribution to achieve descending order) to sort players' contributions. The leaderboard is sorted from highest to lowest contribution, and top-ranked players receive special rewards or privileges (e.g., extra daily usage attempts, higher usage effect bonuses, etc.). Reward and privilege information is stored in server configuration files or a database and displayed to players on the leaderboard interface.

[0061] 3. Improved Friend Assistance Logic and Results (Network Communication - Message Queue, Computation Logic - Linear Interpolation Algorithm) When a player uses an Energy Core and chooses to invite friends for help, the client sends a help invitation request to the server. This request includes the player's character ID and a list of invited friends. Upon receiving the request, the server sends a notification to the invited friends (using push notifications or an in-game messaging system; message queues like RabbitMQ can be used to ensure reliable message delivery and asynchronous processing, placing notification messages in a message queue for individual processing and delivery to friends by a message consumer). Friends can then respond by clicking the "Help" button.

[0062] The server tracks the number of friends who assist the player and calculates the actual improvement in effectiveness based on a pre-set algorithm. For example, for each friend who assists, the power level increase is increased by a certain percentage (e.g., 1%) on top of the base increase, and additional attribute bonuses may be obtained (e.g., a 0.5% increase in critical hit rate). The upper limit of the assistance effect is determined based on the experience orb level and game balance design. A linear interpolation algorithm is used when calculating the assistance effect (e.g., linear interpolation is performed on a pre-set assistance effect curve based on the number of friends assisting to obtain the specific improvement value). The server applies the calculated final effect to the player's character or pet and stores the assistance records (including the list of assisting friends, assistance time, and final effect) in a database for players to query and review.

[0063] In another alternative implementation, a behavioral analysis model is built using a machine learning framework; Features and corresponding labels are defined according to different analysis objectives, and features are extracted periodically from the game data, behavior data and environmental data of the game players collected. The behavior analysis model is trained and updated based on the extracted features and corresponding labels; The corresponding behavioral analysis model is determined based on the type of request received regarding game items; The corresponding behavior analysis model is used to execute corresponding game item operations, which include game item purchase suggestions, game item usage strategies, and game item effects.

[0064] The game system continuously observes and analyzes players' behavioral patterns when using energy cores through machine learning algorithms. For example, the system records when a player chooses to use an experience orb (such as before leveling up a character or facing a specific challenge), the object of use (character or pet preference), and the game decisions made after use (such as whether to immediately participate in combat or challenge a higher-difficulty dungeon). Based on this data, the system continuously adjusts and optimizes its intelligent recommendations and personalized customization schemes for players, making them more aligned with the player's actual gaming style and needs.

[0065] As players progress through the game, the system can predict their need for specific game items (such as energy cores) in different game scenarios and provide corresponding prompts and suggestions in advance. For example, if the system detects that a player has failed to complete a high-difficulty dungeon consecutively, it will proactively prompt the player upon their next login whether they need to use energy cores to improve their strength, and recommend the most suitable usage strategy based on the player's past behavior to help the player better cope with game challenges.

[0066] Furthermore, a game system architecture with self-evolution capabilities should be designed, enabling the energy core system to automatically expand and optimize its functions as the game develops and player needs change. For example, with the introduction of new game content (such as new classes, new dungeons, new gameplay, etc.), the system can automatically identify and adjust the usage rules and effects of energy cores in these new scenarios without requiring large-scale manual updates.

[0067] Player feedback and behavioral data in the game also drive the system's evolution. If a large number of players offer similar suggestions for improvement regarding the use and effects of energy cores in specific game scenarios, the system can automatically make corresponding adjustments and optimizations, such as modifying usage restrictions or adding new attribute bonus types, to improve the overall quality of the game and player satisfaction, and maintain the game's long-term competitiveness and appeal.

[0068] In practice, player behavior can be adaptively learned: 1. Behavioral data collection and storage (log recording - Log4j logging framework, data persistence - data archiving strategy) The game server meticulously records player behavior data throughout the entire process of using Energy Cores. This includes the time players open the Magic Stone Shop, the types of gift packs browsed, the time spent on each pack, and the final purchase decision (purchase or no purchase, and if purchased, the type and quantity of the purchased gift pack); actions related to using Experience Orbs in the inventory (such as the time of use, the target, whether personalization was performed, and the specific content of the customization); and game behavior after using Experience Orbs (such as the type of battle immediately participated in, the battle result, and behavior in PVP or PVE scenarios, such as skill release order, movement path, and interactions with other players or monsters). Log frameworks such as Log4j are used (by configuring parameters such as log levels and output formats) to record this behavioral data, and the log files are stored on the server's local storage. Simultaneously, to facilitate data management and retrieval, a data archiving strategy is employed (such as periodically compressing and archiving old log files by date or size and storing them in a designated archive directory). Log data is periodically organized and archived, stored in a behavioral data analysis table in the database, ensuring data persistence and ease of subsequent querying and analysis.

[0069] 2. Machine Learning Model Training and Updates (Machine Learning Frameworks - Combining TensorFlow and Keras, Algorithm Iteration - Stochastic Gradient Descent Optimization Algorithm) A machine learning framework (such as a combination of TensorFlow and Keras, leveraging Keras' high-level API for rapid model building, with TensorFlow's computing engine providing underlying support) is used to construct the behavioral analysis model. Newly collected player behavior data is extracted from the database periodically (e.g., daily or weekly) to train and update the model. During training, features and labels are defined according to different analytical objectives. For example, using whether a player purchases a specific Energy Core bundle as a label, player browsing behavior data in the store (e.g., browsing time, other related bundles browsed) is used as features to train a classification model (e.g., using a Multilayer Perceptron (MLP) model) to predict player purchase intentions. Using a player's combat performance after using Experience Orbs (e.g., win rate, damage output) as a label, and character status before use, Experience Orb attributes, and environmental factors during use as features, a regression model (e.g., linear regression or neural network regression model) is trained to evaluate the relationship between usage effects and various factors. During model training, stochastic gradient descent (SGD) optimization algorithms and their variants (e.g., Adagrad, Adadelta) are used.

[0070] When customizing game items: 1. Deep Neural Network Model Construction and Training: Collect multi-dimensional data, including detailed records of players' historical experience orbs (such as usage time, experience orb type, changes in character attributes before and after usage, etc.), various data on character growth (level increase curve, skill unlocking and upgrading, frequency of participation in different game modes, etc.), and game goals clearly set by players (obtained through in-game questionnaires or settings options).

[0071] Based on this data, a deep neural network can be built, with the network structure employing a multilayer perceptron (MLP). For example, input layer nodes correspond to various data features, hidden layers perform complex feature transformations using non-linear activation functions (such as ReLU), and the output layer generates recommended weights or priorities for different experience balls. During training, a large amount of player sample data is used, and the network parameters are continuously adjusted through the backpropagation algorithm, enabling the model to accurately capture the experience ball demand patterns of different game styles and target players.

[0072] 2. Real-time adaptive adjustment mechanism: Establish a real-time data monitoring and feedback system to continuously collect new data during the player's gameplay, such as the player's performance in new battle scenarios and feedback on the use of newly acquired experience orbs.

[0073] When a player is detected being frequently defeated by specific enemy skills in a PvP battle, the model quickly analyzes the relevant data and identifies that the player may need experience orbs to improve the corresponding defensive attribute. Based on this, the recommended list for that player is immediately adjusted, prioritizing experience orbs that improve the relevant defensive attribute, and the player is notified via in-game notifications or push messages to ensure that the recommended scheme always closely aligns with the player's dynamic needs in the game.

[0074] In another alternative implementation, when generating the game item adaptation effect: Define a state space and an action space. The state space covers various environmental data of the game process, and the action space includes various possible attribute change strategies of game items. A deep Q-network is used to reinforce the machine learning framework. Through training, the constructed behavioral learning model learns to dynamically formulate the optimal strategy for changing the attributes of game items based on the game environment and the state of the player.

[0075] In practice: 1. Reinforcement learning model design and training: Define a state space that encompasses all environmental factors of the game scene, such as weather conditions (sunny, rainy, snowy, etc.) represented by different numerical codes, terrain features of the map area (plains, mountains, swamps, etc.) and resource distribution, as well as the player's current detailed character status (health, mana, skill cooldown, equipment durability, etc.). The action space includes various possible attribute change strategies for experience orbs, such as increasing the numerical range of different attributes and adjusting the probability of triggering special effects.

[0076] A Deep Q-Network (DQN) or its variants are employed as the reinforcement learning algorithm framework. During training, the agent (i.e., the model) continuously explores action choices under different states in a large number of simulated game environments, updating the Q-value table or policy network based on the player's performance in actual games (e.g., positive rewards for successful task completion and negative rewards for battle failure). After a long period of training, the model learns to dynamically formulate optimal experience ball attribute change strategies based on complex game scenarios and player states.

[0077] 2. Environmental perception and fusion model: Convolutional neural networks (CNNs) are used to process game environment image data to extract visual features such as weather changes, texture and layout features of map areas; at the same time, recurrent neural networks (RNNs) or long short-term memory networks (LSTMs) are used to analyze player behavior sequence data in different environments to capture the correlation between player behavior patterns and the environment.

[0078] These extracted features are then fused with an experience orb attribute calculation model. For example, in a game set in a magical world, when a CNN identifies a high concentration of magical elements in the game scene (such as in a mysterious magical forest area), and an RNN analyzes that the player frequently uses magical skills in that area, the fusion model will increase the bonus ratio of experience orbs to magical damage and mana regeneration attributes based on pre-learned rules, achieving a deep and dynamic adaptation of item attributes to the game environment and player behavior.

[0079] In another alternative implementation, the behavioral data may also include social activity data; The social activity data includes social activity level, social relationship intimacy, and social group goals; The recommendation weight of different game items is dynamically adjusted based on the social activity level. A model of social relationship intimacy among game players is constructed based on the aforementioned social relationship intimacy. The recommendation priority of game items is determined based on the aforementioned social relationship intimacy model; Game items are recommended based on the alignment between the player's game data and the social goals of their social group.

[0080] In practice: 1. Social Activity Indicators Define metrics for a player's social activity, such as the number of times a player participates in activities within social groups and the frequency of interaction with other players. Let A be the social activity metric for player i. i This can be obtained by statistically analyzing relevant data over a certain period of time.

[0081] When recommending experience orbs, the recommendation weight of different experience orbs is adjusted based on the player's social activity level. For example, for players with high social activity, the recommendation weight of team-based experience orbs is increased. Let's assume the base recommendation weight of experience orb j is W. j The adjusted recommendation weight W j' It can be calculated using the following simple formula: W j' = W j + k A i Where k is an adjustment coefficient, which can be determined according to the actual situation. In this way, the experience ball recommendations are more inclined to be types that match the player's social characteristics.

[0082] 2. Social Relationship Intimacy Model Construct a social relationship intimacy matrix between players. If player i and player j frequently team up for games, exchange items, etc., their intimacy level C is calculated. ij High. Intimacy can be quantified by setting rules and behavioral statistics.

[0083] - When player i considers using an experience orb, the system considers the effects and choices made by their close friends using the orb. For example, if most of player i's close friends have made good progress in the game after using a particular experience orb, the system can increase the priority of recommending that experience orb to player i. Let F be the set of player i's close friends. i The average effectiveness of experience orb k among friends is E. k Recommendation priority P ik It can be represented as:

[0084] Where E jk This is an evaluation of the effect of friend j using experience ball k, |F i | represents the size of the friend set. Experience balls are sorted and recommended based on the calculated recommendation priority.

[0085] 3. Social Group Goal-Oriented Recommendation Model Analyze the overall goals and development direction of social groups. For example, one guild focuses on conquering high-difficulty dungeons, while another team focuses on PvP ranking. Let the goal weight vector of social group G be TG = [t1, t2, …,t n ], where t m This indicates the degree of importance attached to the m-th game objective (such as dungeon completion, PVP, etc.).

[0086] For player p in the group, experience orbs are recommended based on the alignment between their individual game data and the group's objectives. For example, if player p performs poorly in game activities related to the group's objectives, the system will recommend experience orbs that help improve their abilities in that area. Let the improvement ability of experience orb l for the m-th objective be U. lm Recommended score: S pl It can be calculated as:

[0087] Experience orbs are selected and recommended based on recommended scores, so that players' selection of experience orbs is more in line with the overall needs and development direction of the team.

[0088] In another alternative implementation, machine learning can be used to learn and analyze players' social interactions: 1. Implementation of Social Network Analysis and Clustering Models: Collect rich data from players on in-game social platforms, including friend lists, chat logs (using natural language processing technology to extract keywords and sentiment), historical records of team participation in activities (such as the number of times they participated in team dungeons, guild wars, etc., contribution level, team ranking, etc.), and detailed information on the gifting and receiving of experience orbs.

[0089] Based on this data, a player social relationship graph is constructed, with nodes representing players and edge weights determined by factors such as interaction frequency and cooperation effectiveness. Graph clustering algorithms (such as the Louvain algorithm) are used to divide players into different social groups, such as closely collaborative, competitive, and casual social groups. Differentiated experience orb recommendation strategies are developed for each group. For closely collaborative groups, experience orb combinations that promote balanced development of team member attributes and improved synergy are recommended, such as those increasing the team's experience sharing ratio and raising the team member's maximum health. For competitive groups, the focus is on recommending experience orbs that quickly enhance individual strength and competitive advantage, such as those increasing critical hit rate and attack speed.

[0090] 2. Construction of prediction models based on social behavior: Using classification and regression algorithms from machine learning, and taking players' social behavior data as input, this study predicts players' potential needs and preferences in social interactions. For example, the decision tree algorithm is used to analyze players' chat content and topics in social groups to determine their level of attention to different game content (such as PVP battles, PVE dungeon exploration, pet raising, etc.); and a linear regression model is used to predict players' demand trends for experience orbs of different attributes based on their experience orb gifting and receiving behavior.

[0091] When the model predicts a player has a high demand for a certain type of experience orb, the system proactively pushes relevant recommendations and provides personalized usage suggestions based on the characteristics and dynamics of the player's social group. For example, if a player's social group is preparing for a large-scale guild dungeon event, and the model predicts that the player needs to improve their defensive attributes, the system will recommend defensive experience orbs suitable for the dungeon's difficulty and the team's needs, and will prompt the player on the appropriate timing and allocation methods for using them within the team. This further enhances the deep integration of social interaction and experience orb recommendations, improving the player's social experience and overall participation in the game.

[0092] In another alternative implementation, such as Figure 2 As shown, a device for operating a game prop includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the method for operating a game prop as described in any of the above embodiments.

[0093] In another alternative embodiment, a computer-readable storage medium stores computer program instructions thereon, which, when executed by a processor, implement the steps of the operation method of a game prop as described in any of the above embodiments.

[0094] In summary, the present invention provides a method, apparatus, and computer-readable storage medium for operating game props. By collecting game data, behavioral data, and environmental data from the game process from game players, and combining them with a machine learning framework, it enables accurate recommendations of game props, personalized customization of game prop usage strategies, and contextualized presentation of game prop effects during the purchase, use, effect presentation, and social interaction of game props. It can achieve personalized recommendations, customization, and flexible presentation based on the specific needs and preferences of game players, thereby improving the flexibility and fun of using game props.

[0095] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent modifications made based on the content of the present invention specification and drawings, or direct or indirect applications in related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for operating a game prop, characterized in that, Including the following steps: Collect game data, behavioral data, and environmental data from game players during gameplay; Receive game item purchase requests and generate game item purchase suggestions based on the game player's game data and behavior data; Receive requests to use game items, and customize and display strategies for using the game items based on the player's behavior data. Receive a request to generate game item effects, and generate appropriate effects based on the environmental data of the game player's current game process.

2. The method for operating a game prop according to claim 1, characterized in that, It also includes the following steps: Receive social group operation requests; Perform the corresponding operation based on the social group operation request; The game item growth value for each player is determined based on the operation results.

3. The method for operating a game prop according to claim 1, characterized in that, Use machine learning frameworks to build behavioral analysis models; Features and corresponding labels are defined according to different analysis objectives, and features are extracted periodically from the game data, behavior data and environmental data of the game players collected. The behavior analysis model is trained and updated based on the extracted features and corresponding labels; The corresponding behavioral analysis model is determined based on the type of request received regarding game items; The corresponding behavior analysis model is used to execute corresponding game item operations, which include game item purchase suggestions, game item usage strategies, and game item effects.

4. A method for operating a game prop according to any one of claims 1 to 3, characterized in that, The game data includes the player's character's growth progress, game level, and pet training status; the behavioral data includes the player's item usage history. The process of generating game item purchase suggestions based on the game player's game data and behavior data includes: The correlation between the items used and the corresponding game character's growth progress, game level, or pet training status is determined based on the player's item usage history. Game item purchase suggestions are generated based on the correlation.

5. A method for operating a game prop according to any one of claims 1 to 3, characterized in that, The behavioral data includes the game player's character development goals; The strategy for customizing and displaying the use of game items based on the game player's behavior data includes: Based on the player's character development goals, determine the game items that match those goals and assign priority or proportion to these items.

6. A method for operating a game prop according to any one of claims 1 to 3, characterized in that, The environmental data includes weather data, map area data, and time data; The adaptation effect generated based on the environmental data of the game player's game process includes: The game level upgrade effect is generated based on the map area data of the game player's current game process; The game item bonus effect is generated based on the weather data during the game. The game item usage rules and effects are generated based on the time data of the player's game process.

7. The method for operating a game prop according to claim 4, characterized in that, The behavioral data also includes social activity data; The social activity data includes social activity level, social relationship intimacy, and social group goals; The recommendation weight of different game items is dynamically adjusted based on the social activity level. A model of social relationship intimacy among game players is constructed based on the aforementioned social relationship intimacy. The recommendation priority of game items is determined based on the aforementioned social relationship intimacy model; Game items are recommended based on the alignment between the player's game data and the social goals of their social group.

8. The method for operating a game prop according to claim 3, characterized in that, When generating effects that adapt to game items: Define a state space and an action space. The state space covers various environmental data of the game process, and the action space includes various possible attribute change strategies of game items. A deep Q-network is used to reinforce the machine learning framework. Through training, the constructed behavioral learning model learns to dynamically formulate the optimal strategy for changing the attributes of game items based on the game environment and the state of the player.

9. An operating device for a game prop, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the operation method of a game prop as described in any one of claims 1 to 8.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the steps of the operation method of a game prop according to any one of claims 1 to 8.

Images