Virtual item recommendation method and device, storage medium and electronic equipment

Abstract

The invention discloses a virtual item recommendation method and device, a storage medium and electronic equipment. The method comprises the steps that first historical consumption information and second historical consumption data of a user account are acquired, the first historical consumption information is used for representing consumption behaviors triggered by the user account in a virtual game, and the second historical consumption information is used for representing the consumption behaviors triggered by the user account in a recent time interval; the consumption behavior is triggered for any one of the at least two virtual props of the virtual game; according to the first historical consumption information, obtaining a price interval preference of the user account in the virtual game; according to the second historical consumption information, obtaining a virtual prop preference of the user account in the virtual game; and obtaining a target virtual item pushed to the user account in combination with the price interval preference and the virtual item preference. The method can be applied to the technical fields of artificial intelligence, big data and the like. The technical problem that the recommendation accuracy of the virtual item is low is solved.

Description

Technical Field

[0001] This application relates to the field of computers, and more specifically, to a method, apparatus, storage medium, and electronic device for recommending virtual props. Background Technology

[0002] In related technologies, in order to improve the gaming experience of players in virtual games, various virtual item consumption activities are often provided to players in virtual games to recommend virtual items for players to purchase and use.

[0003] However, in related technologies, the virtual items recommended to players through the aforementioned virtual game item consumption activities are often of fixed types / fixed prices. This leads to some players being unwilling to purchase the recommended virtual items, failing to meet the diverse needs of different players for virtual items in virtual games, and thus causing the technical problem of low accuracy in virtual item recommendations.

[0004] Therefore, there is a technical problem in the related technologies that the accuracy of virtual item recommendations is relatively low. Summary of the Invention

[0005] This application provides a method, apparatus, storage medium, and electronic device for recommending virtual items, in order to at least solve the technical problem of low accuracy in recommending virtual items in related technologies.

[0006] According to one aspect of the embodiments of this application, a method for recommending virtual items is provided, comprising: obtaining first historical consumption information and second historical consumption data of a user account, wherein the first historical consumption information is used to represent consumption behaviors triggered by the user account in a virtual game, and the second historical consumption information is used to represent consumption behaviors triggered by the user account for any one of at least two virtual items in the virtual game within a recent time interval; obtaining the user account's price range preference in the virtual game based on the first historical consumption information, wherein the price range preference is used to represent the range of resource consumption preferred by the user account when triggering consumption behaviors; obtaining the user account's virtual item preference in the virtual game based on the second historical consumption information, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behaviors for any virtual item; and combining the price range preference and the virtual item preference to obtain a target virtual item to be pushed to the user account, wherein the target virtual item is a virtual item among at least two virtual items that meets the price range preference and the virtual item preference.

[0007] According to another aspect of the embodiments of this application, a virtual item recommendation device is also provided, comprising: a first acquisition unit, configured to acquire first historical consumption information and second historical consumption data of a user account, wherein the first historical consumption information represents consumption behavior triggered by the user account in a virtual game, and the second historical consumption information represents consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time interval; a second acquisition unit, configured to acquire the user account's price range preference in the virtual game based on the first historical consumption information, wherein the price range preference represents the range of resource consumption preferred by the user account when triggering consumption behavior; a third acquisition unit, configured to acquire the user account's virtual item preference in the virtual game based on the second historical consumption information, wherein the virtual item preference represents the user account's preference for triggering consumption behavior for any virtual item; and a fourth acquisition unit, configured to combine the price range preference and the virtual item preference to acquire a target virtual item to be pushed to the user account, wherein the target virtual item is a virtual item among at least two virtual items that meets the price range preference and the virtual item preference.

[0008] As an optional solution, the third acquisition unit includes: a preference analysis module, used to perform preference analysis on the second historical consumption information to obtain the user account's preference degree for each virtual item in any virtual item, wherein the preference degree is used to indicate the probability of the user account purchasing each virtual item; and a first determination module, used to determine the virtual item preference based on the preference degree of each virtual item.

[0009] As an optional solution, the fourth acquisition unit includes: a second determining module, used to determine a first set of virtual items that meet the price range preference from at least two virtual items, and to determine the first virtual item with the highest preference from the first set of virtual items; a third determining module, used to determine the first virtual item as the target virtual item; or, a fourth determining module, used to determine a second set of virtual items that have a preference greater than or equal to a preset preference threshold from at least two virtual items, and to determine the second virtual item that meets the price range and consumes the most resources from the second set of virtual items; and a fifth determining module, used to determine the second virtual item as the target virtual item.

[0010] As an optional solution, the device further includes: a display module, used to display the target item tag corresponding to the target virtual item after obtaining the target virtual item pushed to the user's account by combining price range preference and virtual item preference, wherein the target item tag includes a first item tag and a second item tag, the first item tag is used to represent the item attribute of the target virtual item, and the second item tag is used to represent the amount of resources consumed by the target virtual item.

[0011] As an optional solution, the device further includes: an adjustment module, configured to, after displaying the item tag corresponding to the target virtual item in the first display area, adjust the target item tag corresponding to the target virtual item to the reference item tag corresponding to the reference virtual item in response to an item refresh request, and display the reference item tag, wherein, if the target virtual item is the first virtual item, the reference virtual item is the second virtual item, and if the target virtual item is the second virtual item, the reference virtual item is the first virtual item.

[0012] As an optional solution, the preference analysis module includes: an extraction submodule, used to extract features from the second historical consumption information to obtain user features of the user account in the recent time interval, wherein the user features are used to indicate the user account's preference features for consumption behavior triggered by any virtual item in the recent time interval, and the item features corresponding to any virtual item; and an input submodule, used to input the user features and the item features corresponding to each virtual item into the preference prediction model for preference prediction processing to obtain the preference degree of each virtual item output by the preference prediction model, wherein the preference prediction model is a neural network model obtained by pre-training the model using positive and negative samples.

[0013] As an optional solution, the device further includes: a first acquisition module, configured to acquire a set of historical virtual items recommended to the user account before inputting user characteristics and item characteristics into the preference prediction model, wherein the set of historical virtual items includes a set of third virtual items purchased by the user account and a set of candidate virtual items not purchased by the user account; and a second acquisition module, configured to determine, before inputting user characteristics and item characteristics into the preference prediction model, the candidate virtual item with the highest resource consumption from the set of third virtual items, and determine at least one virtual item from the set of third virtual items whose resource consumption deviates from the resource consumption of the candidate virtual item by less than a first preset deviation threshold. The system comprises: a first positive sample and a second negative sample. The first positive sample is defined as at least one virtual prop whose deviation from the first preset deviation threshold is less than a certain value. The second negative sample is defined as at least one virtual prop whose resource consumption is greater than a certain value from the first preset deviation threshold, before the user features and item features are input into the preference prediction model. The third acquisition module is used to determine, from the candidate virtual prop set, at least one virtual prop whose resource consumption is greater than a certain value from the first preset deviation threshold, and to determine the first negative sample as the first negative sample. The training module is used to train the initial preference prediction model based on the first positive sample and the first negative sample before the user features and item features are input into the preference prediction model, thereby obtaining a trained preference prediction model.

[0014] As an optional solution, the device further includes: a fourth acquisition module, used to acquire a fifth set of virtual items not recommended to the user account before inputting user features and item features into the preference prediction model; a fifth acquisition module, used to identify, before inputting user features and item features into the preference prediction model, multiple virtual items from the fifth set of virtual items whose resource consumption is less than the resource consumption of candidate virtual items and to identify the multiple virtual items with a deviation less than the first preset deviation threshold as second positive samples; a sixth acquisition module, used to identify, before inputting user features and item features into the preference prediction model, multiple virtual items from the fifth set of virtual items whose resource consumption is greater than the resource consumption of candidate virtual items and to identify the multiple virtual items with a deviation greater than the second preset deviation threshold as second negative samples; and a training module, including: a training submodule, used to train the initial preference prediction model using the first positive sample and the second positive sample as positive samples, and the first negative sample and the second negative sample as negative samples, to obtain a trained preference prediction model.

[0015] As an optional solution, the second acquisition unit includes: a data analysis module, used to perform data analysis on the first historical consumption information to obtain a first price range for the amount of resources consumed corresponding to the consumption behavior triggered by the user account in the virtual game, wherein the first price range is used to indicate the preferred amount of resources consumed from the lowest to the highest amount of resources consumed; a sixth determination module, used to determine the first price range as the user account's price range preference in the virtual game when the deviation between the highest and lowest amount of resources consumed is less than a third preset deviation threshold; and a seventh determination module, used to determine the second price range as the user account's price range preference in the virtual game when the deviation between the highest and lowest amount of resources consumed is greater than or equal to the third preset deviation threshold, wherein the second price range is a sub-range of the first price range and is used to indicate the preference from the intermediate amount of resources consumed to the highest amount of resources consumed, wherein the intermediate amount of resources consumed is greater than the lowest amount of resources consumed.

[0016] According to another aspect of the embodiments of this application, a computer program product is provided, the computer program product including a computer program / instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer program / instructions from the computer-readable storage medium, and executes the computer program / instructions, causing the computer device to perform the recommended method for virtual props as described above.

[0017] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the above-described method for recommending virtual items through the computer program.

[0018] In this embodiment, the user account's price range preference in the virtual game is obtained based on the user account's consumption behavior triggered in the virtual game, thus obtaining the user account's consumption habits in the virtual game; and, based on the user account's recent consumption behavior on any of the at least two virtual items in the virtual game, the user account's virtual item preference in the virtual game is obtained, thus obtaining the user account's degree of preference for virtual items in the virtual game. Furthermore, a target virtual item that simultaneously meets the above-mentioned price range preference and virtual item preference is determined from the at least two virtual items and pushed to the user account, achieving the purpose of personalized item recommendations for the user account by combining the user account's price range preference and virtual item preference. The recommended target virtual item satisfies the user account's consumption habits and is an item that the user account prefers to consume, thereby meeting the personalized needs of different players for items in the virtual game, thus achieving the technical effect of improving the accuracy of virtual item recommendations and solving the technical problem of low accuracy in virtual item recommendations in related technologies. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0020] Figure 1 This is a schematic diagram of an application environment for an optional method of recommending virtual props according to an embodiment of this application;

[0021] Figure 2 This is a schematic diagram of the flow of an optional method for recommending virtual props according to an embodiment of this application;

[0022] Figure 3 This is a schematic diagram of an optional method for recommending virtual items according to an embodiment of this application;

[0023] Figure 4 This is a schematic diagram of an optional method for recommending virtual items according to an embodiment of this application;

[0024] Figure 5 This is a schematic diagram of an optional method for recommending virtual items according to an embodiment of this application;

[0025] Figure 6This is a schematic diagram of an optional method for recommending virtual items according to an embodiment of this application;

[0026] Figure 7 This is a schematic diagram of an optional virtual prop recommendation device according to an embodiment of this application;

[0027] Figure 8 A schematic diagram of the structure of an optional electronic device according to an embodiment of this application. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] According to one aspect of the embodiments of this application, a method for recommending virtual items is provided. Optionally, as an optional implementation, the above-described method for recommending virtual items may be applied to, but is not limited to, [examples of other methods]. Figure 1 The environment shown may include, but is not limited to, a client 102 and a server 112. The client 102 may include, but is not limited to, a display 104, a processor 106, and a memory 108. The server 112 includes a database 114 and a processing engine 116.

[0031] The specific process can be summarized in the following steps:

[0032] In step S102, the client 102 obtains a virtual item recommendation request triggered on the user account. The virtual item recommendation request is triggered by touching a button preset in the virtual game to expand and display the item activity page. The virtual item recommendation request is used to request the push of target virtual items to the user account.

[0033] In steps S104-S106, client 102 sends a virtual item recommendation request to server 112;

[0034] In step S108, server 112 responds to the virtual item recommendation request and obtains the first historical consumption information and the second historical consumption data of the user account. The first historical consumption information is used to represent the consumption behavior triggered by the user account in the virtual game, and the second historical consumption information is used to represent the consumption behavior triggered by the user account for any one of the at least two virtual items in the virtual game within a recent time period.

[0035] Step S110: Based on the first historical consumption information, obtain the user account's price range preference in the virtual game, wherein the price range preference is used to represent the range of resource consumption when the user account triggers a consumption behavior; and based on the second historical consumption information, obtain the user account's virtual item preference in the virtual game, wherein the virtual item preference is used to represent the user account's preference for triggering a consumption behavior for any virtual item.

[0036] Step S112: Combining price range preference and virtual item preference, determine the target virtual item that matches the price range preference and virtual item preference from at least two virtual items;

[0037] Steps S114-S116: The target virtual item is pushed to the client 102 via the network 110. The processor 106 in the client 102 is used to receive and process the target virtual item and its related data, display the target virtual item on the display 104, and store the related data in the memory 108.

[0038] remove Figure 1 Beyond the examples shown, the above steps can be completed independently by the client or server, or collaboratively by both, such as by client 102 executing steps S108 to S112, thereby reducing the processing load on server 112. Client 102 includes, but is not limited to, laptops, tablets, desktop computers, smart TVs, etc., and this application does not limit the specific implementation of client 102. Server 112 can be a single server, a server cluster consisting of multiple servers, or a cloud server.

[0039] Alternatively, as an alternative implementation method, such as Figure 2As shown, the recommended methods for using virtual items can be performed by electronic devices, such as... Figure 1 The client or server shown includes the following specific steps:

[0040] S202, obtain the first historical consumption information and the second historical consumption data of the user account, wherein the first historical consumption information is used to represent the consumption behavior triggered by the user account in the virtual game, and the second historical consumption information is used to represent the consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time period.

[0041] S204, Based on the first historical consumption information, obtain the user account's price range preference in the virtual game, wherein the price range preference is used to represent the range of resources preferred to be consumed when the user account triggers a consumption behavior;

[0042] S206, Based on the second historical consumption information, obtain the user account's virtual item preference in the virtual game, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behavior for any virtual item;

[0043] S208, combining price range preference and virtual item preference, obtain the target virtual item to be pushed to the user's account, wherein the target virtual item is a virtual item that meets the price range preference and virtual item preference among at least two virtual items.

[0044] Optionally, in this embodiment, the above-mentioned method for recommending virtual items can be applied, but is not limited to, in the scenario of recommending game gift packs. In this scenario, only a limited number of game gift packs are selected from multiple available game gift packs and displayed to the user's account to enhance the user's purchase feedback on the displayed game gift packs.

[0045] It should be noted that, regarding the selection of a limited number of game gift packs, the relevant technology often involves selecting multiple identical game gift packs of a fixed type / fixed price and recommending them to various user accounts. This means that different user accounts can purchase the same virtual items in the game. However, this method cannot meet the diverse needs of different players for virtual items in the game; in other words, it suffers from low accuracy in recommending virtual items.

[0046] To address the aforementioned shortcomings, the above-mentioned virtual item recommendation method combines the user account's price range preferences and virtual item preferences to provide personalized item recommendations. The recommended target virtual items match the user account's consumption habits and are items that the user account prefers to consume. This can satisfy the personalized needs of different players for items in virtual games, thereby improving the technical effect of virtual item recommendation accuracy and solving the aforementioned shortcoming of low virtual item recommendation accuracy.

[0047] Optionally, in this embodiment, the virtual items in the virtual game may be, but are not limited to, gift pack-type items in the virtual game, which are displayed to the user account. The user account can purchase and use them by consuming a certain amount of resources (such as virtual gold coins).

[0048] It should be noted that among the virtual items mentioned above, different virtual items have a natural resource order. That is, for different virtual items, the amount of resources a user account needs to consume to acquire them can be different, and generally, the more resources required, the higher the item attributes (such as item quantity, item priority, etc.) given by the corresponding virtual item.

[0049] It's also important to note that different user accounts exhibit different spending habits. However, relying solely on a user account's spending history cannot guarantee a true reflection of their actual spending preferences for virtual items. For example, a user account might intend to purchase a high-value virtual item in a game (due to its discounted nature, this purchase offers a better deal than other channels). If the game only recommends and displays a lower-value bundle, the user account might buy the lower-value bundle due to the discount factor. However, this purchase does not reflect the user account's true spending intentions. In other words, a user account's spending history is often inaccurate and may not accurately reflect their genuine spending preferences for virtual items.

[0050] To address the aforementioned shortcomings, this embodiment determines a user account's price range preference based on first historical consumption information of the user account's spending in the virtual game, and determines the user account's virtual item preference based on second historical consumption information of the user account's spending on at least two virtual items in the virtual game within a recent time period. Then, by combining the price range preference and the virtual item preference, the target virtual items to be pushed to the user account are determined. Thus, by estimating the user account's spending capacity range based on the user account's long-term historical consumption characteristics in the virtual game, and further determining the range of resource consumption for the user account's preferred virtual items, and by estimating the user account's consumption preference for each virtual item based on the user account's recent historical consumption characteristics in the virtual game, comprehensive and multi-dimensional information can be combined to personalized determine the virtual items that each user account wishes to be recommended and displayed, reflecting the user account's true consumption preferences for virtual items.

[0051] Optionally, in this embodiment, the first historical consumption information of the user account is used to represent the consumption behavior triggered by the user account in the virtual game, wherein the consumption behavior triggered by the user account in the virtual game includes any consumption behavior made by the user account in the virtual game.

[0052] Optionally, in this embodiment, the second historical consumption information of the user account is used to represent the consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time period, wherein the at least two virtual items are virtual items in the virtual game to be displayed to the user account for consumption.

[0053] It should be noted that in virtual games, user accounts can trigger consumption behaviors that include, but are not limited to, purchasing virtual items from at least two virtual items. In addition, it can also include, but is not limited to, other types of consumption behaviors, such as recharging resources for other user accounts.

[0054] Understandably, the first historical consumption information is used to characterize the user's account's consumption habits in the virtual game, which is a long-term, global feature, while the second historical consumption information is used to characterize the account's recent consumption habits of virtual items in the virtual game, which is a short-term, local feature.

[0055] Optionally, in this embodiment, the user account's price range preference in the virtual game is obtained based on the first historical consumption information, wherein the price range preference is used to represent the range of resources that the user account prefers to consume when triggering consumption behavior in the virtual game.

[0056] Optionally, in this embodiment, price range preference can be represented, but is not limited to, in the form of discrete or continuous intervals.

[0057] To further illustrate, if the first historical consumption information indicates that a user's account has made 5 consumption behaviors in the virtual game, with the corresponding resource consumption amounts being 5, 10, 10, 5, and 15 respectively, then the price range preference can be determined as a discrete interval (set) {5, 10, 15}, or as a continuous interval [5, 15].

[0058] It should be noted that, in order to further improve the accuracy of predicting user account consumption habits, after obtaining the user's historical consumption behavior data, the consumption behavior data can be filtered and / or amplified to reduce the noise in the user account's consumption records. Then, based on the amount of resources consumed indicated by the filtered and / or amplified consumption behavior data, the corresponding discrete or continuous intervals can be determined, thereby obtaining the user account's price range preference.

[0059] Optionally, in this embodiment, the user account's virtual item preference in the virtual game is obtained based on the second historical consumption information, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behavior on any of the virtual items in a recent time period.

[0060] Optionally, in this embodiment, virtual item preferences can be represented by the user account's preference for each virtual item, but are not limited to the preference degree of each virtual item, wherein the preference degree is used to indicate the probability of the user account purchasing each virtual item.

[0061] It should be noted that obtaining the user's preference for each virtual item can be done, but is not limited to, using a pre-trained preference prediction model to output the aforementioned preference. The preference prediction model is a neural network model obtained by pre-training the model using positive and negative samples. It is used to perform preference analysis (preference prediction) on the relevant features of the input second historical consumption information to output the user account's preference for the virtual items involved in the second historical consumption information.

[0062] The embodiments provided in this application obtain the user account's price range preference in the virtual game based on the user account's consumption behavior triggered in the virtual game, thereby obtaining the user account's consumption habits in the virtual game; and obtain the user account's virtual item preference in the virtual game based on the user account's consumption behavior triggered on any of the at least two virtual items in the virtual game in the recent period, thereby obtaining the user account's degree of preference for virtual items in the virtual game. Furthermore, a target virtual item that simultaneously meets the above-mentioned price range preference and virtual item preference is determined from at least two virtual items and pushed to the user account, achieving the purpose of providing personalized item recommendations for the user account by combining the user account's price range preference and virtual item preference. The recommended target virtual item satisfies the user account's consumption habits and is an item that the user account prefers to consume, thereby meeting the personalized needs of different players for items in the virtual game and achieving the technical effect of improving the accuracy of virtual item recommendations.

[0063] As an optional approach, based on second historical consumption information, the user account's virtual item preferences in the virtual game can be obtained, including:

[0064] S1, perform preference analysis on the second historical consumption information to obtain the user account's preference degree for each virtual item in any virtual item, where the preference degree is used to indicate the probability that the user account will purchase each virtual item;

[0065] S2, determine virtual item preferences based on the preference level of each virtual item.

[0066] Optionally, in this embodiment, performing preference analysis on the second historical consumption information may include, but is not limited to, using a pre-trained preference prediction model to predict the preferences of the relevant features of the input second historical consumption information and output the aforementioned preference degree, wherein the preference prediction model is a neural network model obtained by pre-training the model using positive and negative samples.

[0067] Optionally, in this embodiment, the preference degree is used to indicate the probability that a user account will purchase each virtual item. The value is between 0 and 1. The higher the preference degree, the higher the probability that the user account will purchase the virtual item (i.e., trigger the consumption behavior).

[0068] It should be noted that virtual item preferences are determined based on the preference level of each virtual item. It is understood that virtual item preferences can, but are not limited to, indicating a user account's preference level for each virtual item, and can also, but are not limited to, indicating multiple preferred virtual items whose preference level, derived from the preference levels of the aforementioned virtual items, exceeds a preset threshold.

[0069] The embodiments provided in this application determine a user account's virtual item preferences based on the user account's preference for various virtual items. These preferences are then used to make personalized item recommendations for the user account in combination with the user account's price range preferences. This can meet the personalized needs of different players for items in virtual games, thereby improving the technical effect of recommending virtual items.

[0070] As an optional approach, by combining price range preferences and virtual item preferences, the target virtual items to be pushed to the user's account can be obtained, including:

[0071] S1, determine a first set of virtual items that meets the price range preference from at least two virtual items, and determine the first virtual item with the highest preference from the first set of virtual items;

[0072] S2, designate the first virtual item as the target virtual item; or...

[0073] S3, determine a second set of virtual props from at least two virtual props whose preference degree is greater than or equal to a preset preference threshold, and determine the second virtual prop that consumes the most resources from the second set of virtual props;

[0074] S4, designate the second virtual item as the target virtual item.

[0075] Optionally, in this embodiment, each of the at least two virtual items corresponds to a quantity of resources required to acquire / purchase, and also corresponds to a preference degree, thereby determining a first set of virtual items that conforms to the price range preference, wherein the quantity of resources required to acquire / purchase the virtual items in the first set of virtual items belongs to the range of the quantity of resources consumed by the preference indicated by the price range preference.

[0076] It should be noted that after determining the first set of virtual items, the first virtual item with the highest preference is selected from it and designated as the target virtual item to be pushed to the user's account.

[0077] It should be noted that if there are at least two virtual items with the highest preference and the same value in the first set of virtual items, the virtual item with the lower amount of resources required to obtain / purchase it will be identified as the target virtual item.

[0078] Understandably, the process involves first filtering from at least two virtual items based on the user account's spending power indicated by the price range preference, to obtain the first set of virtual items that the user account can purchase. Then, based on the user account's preference information for virtual items indicated by the preference degree, the target virtual item that the user account can purchase and most desires to purchase is determined from the first set of virtual items.

[0079] Optionally, in this embodiment, a second set of virtual props may be determined from at least two virtual props, wherein the preference degree corresponding to the virtual props in the second set of virtual props is greater than or equal to a preset preference threshold.

[0080] It should be noted that after determining the second set of virtual items, the second virtual item that requires the highest amount of resources to purchase and falls within the price range is selected and designated as the target virtual item to be pushed to the user's account.

[0081] Understandably, the higher the resource consumption of a virtual item, the better the attributes it brings to the user account. For example, in the case of a discount-type gift pack, the higher the resource consumption, the greater the discount. Therefore, while ensuring that the user account has a high preference for virtual items and is able to purchase them, the virtual items that consume the highest amount of resources are further identified as the target virtual items to be pushed to users.

[0082] It should be noted that if there are at least two virtual items in the second set that meet the price range and consume the highest (and the same) amount of resources, the virtual item with higher preference will be identified as the target virtual item.

[0083] Understandably, the process involves first filtering from at least two virtual items based on the user account's preference information for virtual items indicated by the preference level, to obtain a second set of virtual items that the user account is more likely to purchase. Then, based on the user account's spending power indicated by the price range preference, the target virtual item that the user account is most likely to purchase and that consumes the most resources is determined from the first set of virtual items.

[0084] The embodiments provided in this application combine price range preference and preference degree to perform two separate screenings of at least two virtual items under different conditions, thereby determining the target virtual item to be pushed to the user's account. Different screening strategies can be adopted according to actual needs, that is, first screening is performed based on one of the conditions of price range preference and preference degree, and then the results of the first screening are searched according to the other condition to obtain the target virtual item. This not only meets the personalized needs of different players for items in virtual games, but also improves the flexibility and accuracy of virtual item recommendations.

[0085] As an alternative approach, after combining price range preferences and virtual item preferences to obtain the target virtual items to be pushed to the user's account, the method also includes:

[0086] S1, Display the target item tag corresponding to the target virtual item, wherein the target item tag includes a first item tag and a second item tag. The first item tag is used to represent the item attribute of the target virtual item, and the second item tag is used to represent the amount of resources consumed by the target virtual item.

[0087] Optionally, in this embodiment, the first item tag is used to represent the item attributes of the target virtual item, and may include, but is not limited to, at least one of the following: quantity information of the target virtual item, and attribute information of the target virtual item, wherein the quantity information is used to represent the number of items included in the target virtual item, and the attribute information is used to represent the beneficial effect that the target virtual item can provide in the virtual game. The second item tag is used to represent the amount of resources required to purchase the virtual item.

[0088] To further illustrate, such as Figure 3 As shown, the client displaying the virtual game shows two virtual items (virtual item 302 and virtual item 308), along with their corresponding item tags. Virtual item 302 is a fixed-type virtual item, meaning it's the same for all user accounts. Virtual item 308 is a non-fixed-type virtual item, the target virtual item recommended to the user account mentioned above; different user accounts may have different virtual items.

[0089] Specifically, the first item label 304 is used to indicate the item attributes of the virtual item 302, such as "100 virtual coins + 20 virtual coins (bonus)"; the second item label 306 is used to indicate the amount of resources consumed by the virtual item 302, such as "10 resources consumed"; the second item label 310 is used to indicate the item attributes of the virtual item 308, such as "200 virtual coins + 60 virtual coins (bonus)"; and the second item label 312 is used to indicate the amount of resources consumed by the virtual item 308, such as "20 resources consumed".

[0090] The embodiments provided in this application display the item attributes and resource consumption of the target virtual item pushed to the user's account through a first item tag and a second item tag, enabling the user to quickly understand the information of the recommended virtual item, thereby improving the user's virtual item acquisition experience.

[0091] As an optional approach, after displaying the item tag corresponding to the target virtual item in the first display area, the method further includes:

[0092] S1, in response to the item refresh request, adjust the target item tag corresponding to the target virtual item to the reference item tag corresponding to the reference virtual item, and display the reference item tag. Wherein, if the target virtual item is the first virtual item, the reference virtual item is the second virtual item, and if the target virtual item is the second virtual item, the reference virtual item is the first virtual item.

[0093] Optionally, in this embodiment, the item refresh request may be, but is not limited to, triggered by touching the displayed item refresh label, and is used to request an update of the target virtual item and its corresponding target item label. The item refresh label may be, but is not limited to, displayed together with the first item label and the second item label on the display interface of the client running the virtual game.

[0094] Optionally, in this embodiment, in response to the item refresh request, the target virtual item is updated, the target item tag corresponding to the target virtual item is updated, and the updated item tag is displayed.

[0095] Understandably, when the target virtual item is the first virtual item, in response to the item refresh request, the target virtual item is updated to the second virtual item to obtain a reference virtual item, and the target item tag corresponding to the target virtual item is updated to the reference virtual tag corresponding to the second virtual item. The reference virtual tag includes a first reference item tag and a second reference item tag. The first reference item tag is used to describe the item attributes of the reference virtual item, and the second reference item tag is used to indicate the amount of resources consumed by the reference virtual item.

[0096] It should be noted that acquiring the first virtual item may include, but is not limited to, determining a set of first virtual items that match the price range preference from at least two virtual items, and determining the first virtual item with the highest preference from the first set of virtual items. Acquiring the second virtual item may include, but is not limited to, determining a set of second virtual items with a preference greater than or equal to a preset preference threshold from at least two virtual items, and determining the second virtual item that matches the price range and consumes the most resources from the second set of virtual items.

[0097] Optionally, in this embodiment, when the first virtual item and the second virtual item are different virtual items, the target virtual item and the reference virtual item are one of the first virtual item and the other of the second virtual item, respectively.

[0098] It should be noted that when the first virtual item and the second virtual item are the same virtual item, the reference virtual item may be, but is not limited to, the third virtual item. The third virtual item may be, but is not limited to, any one of the following: the virtual item with the highest preference in the first set of virtual items (excluding the first virtual item), or the virtual item in the second set of virtual items (excluding the second virtual item) that meets the price range and consumes the most resources.

[0099] It should be noted that triggering an item refresh request indicates that the user account is dissatisfied with the current target virtual item and its corresponding resource consumption, which may be too high or too low. In this case, a reference virtual item can be set as a candidate virtual item. The resource consumption of the candidate virtual item differs from that of the target virtual item by a predetermined degree. This predetermined degree of difference means that the ratio obtained by dividing the absolute value of the difference between the resource consumption of the candidate virtual item and the target virtual item by the resource consumption of the target virtual item is greater than a first predetermined ratio and less than a second predetermined ratio. This ensures that the reference virtual item and the target virtual item have a certain difference in resource consumption, neither too low nor too high, thereby improving the fault tolerance of virtual game item recommendations.

[0100] In the embodiments provided in this application, the displayed target virtual item and the refreshed reference virtual item are one of the first virtual item and the other of the second virtual item, respectively. Therefore, when the first virtual item fails to satisfy the user and trigger consumption behavior, the system refreshes and recommends the second virtual item to the user, and vice versa, thereby improving the fault tolerance and accuracy of virtual item recommendations.

[0101] As an optional approach, preference analysis is performed on the second historical consumption information to obtain the user account's preference for each virtual item among all virtual items, including:

[0102] S1, extract features from the second historical consumption information to obtain the user features of the user account in the recent time interval. The user features are used to indicate the user account's preference features for consumption behavior triggered by the third virtual item in the recent time interval, and the item features corresponding to the third virtual item.

[0103] S2, input the user features and the item features corresponding to each virtual prop into the preference prediction model, perform preference prediction processing, and obtain the preference degree of each virtual prop output by the preference prediction model. The preference prediction model is a neural network model obtained by pre-training the model using positive and negative samples.

[0104] Optionally, in this embodiment, feature extraction is performed on the second historical consumption information to obtain the aforementioned user features. The user features are used to represent the user account's consumption habits of virtual items in a recent time period, including the preference features of the consumption behavior of the virtual items and the item features of the corresponding virtual items.

[0105] It is understood that the aforementioned consumer behavior preference characteristics may, but are not limited to, reflect the user account's preference for the amount of resources consumed by virtual items, and the aforementioned item characteristics may, but are not limited to, reflect the item attributes corresponding to the virtual items (such as item quantity, item bonus attributes, etc.).

[0106] Optionally, in this embodiment, a pre-trained preference prediction model is used to perform preference prediction processing on the input user features and the item features corresponding to each virtual prop, and output the preference degree corresponding to each virtual prop.

[0107] The embodiments provided in this application use a pre-trained preference prediction model to perform preference prediction processing on the user features obtained from feature extraction and the item features corresponding to each virtual item, thereby obtaining the preference degree of each virtual item. This allows for the rapid and efficient acquisition of the user account's preference degree for each virtual item, thus improving the recommendation efficiency of virtual items.

[0108] As an alternative approach, before inputting user and item features into the preference prediction model, the method also includes:

[0109] S1, retrieve the set of historical virtual items that have been recommended to the user account, where the set of historical virtual items includes the set of third virtual items that the user account has purchased and the set of candidate virtual items that the user account has not purchased;

[0110] S2, determine the candidate virtual item with the highest resource consumption from the third set of virtual items, and determine at least one virtual item from the third set of virtual items whose resource consumption is less than the resource consumption of the candidate virtual item by a degree less than a first preset deviation threshold, and determine at least one virtual item whose deviation is less than the first preset deviation threshold as the first positive sample.

[0111] S3, determine at least one virtual item from the candidate virtual item set whose resource consumption quantity deviates from the resource consumption quantity of the candidate virtual item by more than a second preset deviation threshold, and determine at least one virtual item whose deviation quantity is greater than the second preset deviation threshold as the first negative sample.

[0112] S4. Based on the first positive sample and the first negative sample, train the initial preference prediction model to obtain the trained preference prediction model.

[0113] Optionally, in this embodiment, the third set of virtual items includes multiple virtual items that have been recommended to the user account and purchased by the user account, while the candidate set of virtual items includes multiple virtual items that have been recommended to the user account and not purchased by the user account.

[0114] It should be noted that a user account may have purchased multiple virtual items at different prices in the past. Although some virtual items may have lower resource consumption, meaning a smaller discount, they are still discounted, so the user account may still make the purchase. This part of the historical consumption data should be excluded and not used as a positive sample reflecting the user account's true consumption intentions.

[0115] Specifically, the candidate virtual item with the highest resource consumption is determined from the third set of virtual items, and at least one virtual item whose resource consumption deviates from that of the candidate virtual item by less than a first preset deviation threshold is determined from the third set of virtual items. The at least one virtual item with a deviation less than the first preset deviation threshold is determined as the first positive sample.

[0116] Understandably, if the deviation between the amount of resources consumed and the amount of resources consumed by the candidate virtual item is greater than or equal to the first preset deviation threshold, it indicates that the amount of resources consumed by the virtual item is relatively small, and it is often purchased by the user account because the user does not want to waste the discount. Therefore, it is not considered a positive sample reflecting the user account's true consumption intention.

[0117] It should also be noted that for virtual items that a user account has not purchased in its history, there may be virtual items that some user accounts would be willing to buy, but the discount may not be as good as the target virtual item. This part of the historical consumption data should be excluded and cannot be used as negative sample data reflecting the user account's true consumption intention.

[0118] Specifically, at least one virtual item is identified from the candidate virtual item set whose resource consumption deviates from the resource consumption of the candidate virtual item by a degree greater than a second preset deviation threshold, and the at least one virtual item whose deviation deviates from the second preset deviation threshold is identified as the first negative sample.

[0119] Understandably, if the deviation between the amount of resources consumed and the amount of resources consumed by the candidate virtual item is less than or equal to the second preset deviation threshold, it indicates that the amount of resources consumed by the virtual item is relatively large. This is often because there are target virtual items with better discounts available, and the item was not purchased. Therefore, it is not considered a negative sample reflecting the user's true consumption intention.

[0120] It should be noted that the initial preference prediction model is trained based on the first positive sample and the first negative sample to obtain the trained preference prediction model.

[0121] The embodiments provided in this application divide the historical virtual item set recommended to user accounts into a third set of virtual items already purchased by the user account and a set of candidate virtual items not yet purchased. Based on the third set of virtual items, positive samples are filtered to obtain a first positive sample, and based on the candidate set of virtual items, negative samples are filtered to obtain a first negative sample. The model is then trained using the first positive and first negative samples to obtain the trained preference prediction model. By filtering and processing the sample data, information that more realistically reflects the user account's willingness to consume virtual items is obtained, thereby improving the training quality of the preference prediction model. This achieves the technical effect of improving the accuracy of the output of preferences for each virtual item, and also improves the overall accuracy of virtual item recommendations.

[0122] As an alternative approach, before inputting user and item features into the preference prediction model, the method also includes:

[0123] S1, retrieve the fifth set of virtual items that were not recommended to the user's account;

[0124] S2, determine from the fifth set of virtual props that the deviation between the amount of resources consumed and the amount of resources consumed by the candidate virtual props is less than the first preset deviation threshold, and determine the virtual props with the deviation less than the first preset deviation threshold as the second positive sample;

[0125] S3, identify multiple virtual items from the fifth set of virtual items whose resource consumption is greater than the resource consumption of the candidate virtual items, and determine these multiple virtual items whose deviation is greater than the second preset deviation threshold as the second negative samples.

[0126] Based on the first positive sample and the first negative sample, the initial preference prediction model is trained to obtain a trained preference prediction model, including:

[0127] S4. Using the first positive sample and the second positive sample as positive samples, and the first negative sample and the second negative sample as negative samples, train the initial preference prediction model to obtain the trained preference prediction model.

[0128] Optionally, in this embodiment, the fifth set of virtual items includes multiple virtual items that have not been recommended to the user's account.

[0129] It should be noted that among the virtual items not recommended to user accounts, there may be items that users prefer to purchase, and there may be items that users prefer not to purchase. Therefore, it is necessary to combine the aforementioned first preset deviation threshold, the aforementioned second preset deviation threshold, and the resource consumption amount of the aforementioned candidate virtual items to filter out positive and negative samples reflecting the user account's true consumption intentions from the virtual items not recommended to the user account.

[0130] Specifically, from the fifth set of virtual props, a number of virtual props whose resource consumption is less than the resource consumption of the candidate virtual props are determined, and the number of virtual props whose deviation is less than the first preset deviation threshold is determined as the second positive sample.

[0131] It is understandable that if the deviation between the amount of resources consumed and the amount of resources consumed by the candidate virtual item is less than the first preset deviation threshold, it indicates that the amount of resources consumed by the virtual item is relatively large, which is often what the user account is willing to buy, and serves as a positive sample reflecting the user account's true consumption intention.

[0132] In addition, from the fifth set of virtual items, a number of virtual items whose resource consumption deviates from the resource consumption of the candidate virtual items by a greater degree than a second preset deviation threshold are identified, and the number of virtual items whose deviation exceeds the second preset deviation threshold is identified as the second negative sample.

[0133] It is understandable that if the deviation between the amount of resources consumed and the amount of resources consumed by the candidate virtual item is greater than the second preset deviation threshold, it indicates that the amount of resources consumed by the virtual item is relatively small, and it is often something that user accounts are unwilling to buy, thus serving as a negative sample reflecting the user account's true consumption intention.

[0134] It should be noted that the first positive sample and the second positive sample are used as positive samples, and the first negative sample and the second negative sample are used as negative samples to train the initial preference prediction model, thus obtaining the trained preference prediction model.

[0135] Through the embodiments provided in this application, positive and negative samples are filtered from a fifth set of virtual items that have never been recommended to a user's account to obtain a second positive sample and a second negative sample. The model is then trained based on the first positive sample, the second positive sample, the first negative sample, and the second negative sample to obtain the trained preference prediction model. By filtering and processing the sample data, information that more realistically reflects a user's willingness to consume virtual items is obtained, thereby improving the training quality of the preference prediction model. This achieves the technical effect of improving the accuracy of the output of preferences for each virtual item, and also improves the overall accuracy of virtual item recommendations.

[0136] As an optional approach, based on initial historical consumption information, the user account's price range preference in virtual games can be obtained, including:

[0137] S1, perform data analysis on the first historical consumption information to obtain the first price range of the amount of resources consumed corresponding to the consumption behavior triggered by the user account in the virtual game, wherein the first price range is used to indicate the preferred amount of resources consumed from the lowest to the highest amount of resources consumed.

[0138] S2, if the deviation between the highest and lowest resource consumption is less than the third preset deviation threshold, the first price range is determined as the user account's price range preference in the virtual game.

[0139] S3, when the deviation between the highest and lowest resource consumption is greater than or equal to the third preset deviation threshold, the second price range is determined as the user account's price range preference in the virtual game. The second price range is a sub-range of the first price range and is used to indicate the preference from the middle resource consumption to the highest resource consumption, where the middle resource consumption is greater than the lowest resource consumption.

[0140] Optionally, in this embodiment, the first historical consumption information can be analyzed to obtain a first price range of resource consumption corresponding to the consumption behavior triggered by the user account in the virtual game. The first price range is used to indicate the preferred resource consumption from the lowest to the highest resource consumption.

[0141] It should be noted that the first price range can be, but is not limited to, a discrete range, such as {5, 10, 15}, or it can be, but is not limited to, a continuous range, such as [5, 15].

[0142] It should be noted that, in this embodiment, the first historical consumption information (or features extracted based on the first historical consumption information) can also be used as input to the price range prediction model to perform price range prediction processing and obtain the first price range output by the price range prediction model. The price range prediction model is a pre-trained neural network model used to analyze and predict the first historical consumption information and output a price range to indicate the user account's spending power.

[0143] Optionally, in this embodiment, if the deviation between the highest and lowest resource consumption indicated by the first price range is less than a third preset deviation threshold, it indicates that the length of the first price range is appropriate, and the first price range is directly determined as the user account's price range preference in the virtual game.

[0144] It should be noted that if the deviation between the highest and lowest resource consumption is greater than or equal to the third preset deviation threshold, it indicates that the length of the first price range is too large. Furthermore, considering that generally, the higher the resource consumption of virtual items, the better the corresponding discount or game attributes, and the more popular they are with users within their spending power, when the length of the first price range is too large, in order to more accurately recommend virtual items while also ensuring users' willingness to spend, the earlier partial price ranges (sub-ranges) of the first price range are used as the aforementioned price range preference.

[0145] Through the embodiments provided in this application, when the price range of a user account is large, in order to improve the benefit obtained from virtual items, a sub-range with a higher price range is selected as the price range preference. This realizes the recommendation of high-quality virtual game items under the premise that the user's spending power allows, thereby improving the user's gaming experience.

[0146] As an alternative, the aforementioned method for recommending virtual items can be applied to the scenario of recommending game discount bundles. In this scenario, game discount bundles naturally exhibit a price ranking attribute among different bundles. Furthermore, the relevant technology for recommending game discount bundles has the following two shortcomings:

[0147] Flaw 1): User feedback is noisy and may not reflect users' true preferences. For example, a high-value user might need to purchase a high-priced gift package (due to its discount, this purchase offers a better deal than other channels). If the platform only displays low-priced gift packages, the user might buy the low-priced package because of the discount. However, this purchase behavior does not reflect the user's true intentions. The user most wants to see the high-priced gift package, but from a purchase feedback perspective, the user is giving positive feedback on the low-priced gift package.

[0148] Defect 2): Sparse and unbalanced user purchase data. In recommendation scenarios, purchase behavior is very sparse compared to exposure. Furthermore, in this scenario, user purchase data follows the Pareto principle (80 / 20 rule), meaning 80% of users will buy low-priced bundles, and only 20% will buy high-priced bundles. However, the 20% who buy high-priced bundles account for 80% of revenue. Therefore, historical data suffers from an imbalance between positive and negative samples, and an imbalance between different items within the positive sample.

[0149] To overcome the aforementioned shortcomings, this embodiment proposes a dual-stream mechanism of coarse and fine models to recommend gift packs, based on the aforementioned method for recommending virtual items. The coarse model utilizes the user's long-term historical characteristics (multiple periods of purchase and consumption data in the virtual game) to estimate the user's purchasing power range and determine the range of gift packs the user can purchase. The fine model uses data from the most recent activity (the most recent period of purchase and consumption data of virtual items in the virtual game) to construct positive and negative samples and estimate the user's preference for each gift pack.

[0150] Furthermore, in order to filter noise and alleviate sample imbalance, this embodiment uses a data-centric artificial intelligence approach and proposes a data filtering and augmentation mechanism for gift pack type items.

[0151] To further illustrate, such as Figure 4 As shown, the user's long-term historical features and the item features of the gift pack are input into the coarse model 402 to obtain the user's gift pack range prediction information. After filtering and augmenting the recent user features and the item features of the gift pack, the data is input into the fine model 404 to obtain the user's gift pack preference information. Then, by combining the user's gift pack range prediction information and the user's gift pack preference information, the final recommendation result of the gift pack is determined.

[0152] Specifically, the data filtering section above filters out positive samples where users might be buying simply to avoid discount losses due to exposure to low prices. Assuming users' preferences for items don't change in the short term (e.g., for a week-long event), to maximize user spending, the highest price of a gift pack is chosen as the anchor point. If a gift pack price deviates from the highest price by more than a certain threshold, this type of positive sample is filtered out. (In practice, gift pack tiers can be sorted in ascending order of price, with the threshold set at a difference of more than 3 between two tiers). Under this mechanism, users' positive sample preferences converge, avoiding the appearance of identical or similar user characteristics, such as different labels for the same item and similar labels for items with significantly different prices. For example, if a user purchases a gift pack of value A on the first day of the event and a gift pack of value B on the fourth day, the same user with similar characteristics might prefer both significantly different gift packs. However, the user's preference for gift pack B doesn't accurately reflect their actual preference for gift pack A.

[0153] The data augmentation described above is used to optimize negative samples in the training data. A common practice is to select data from users' exposure-based purchase behavior that were not purchased on a given day as negative samples. Since the gift packages are similar in attributes, differing only in price, it can be assumed that if a user prefers a gift package at a certain price, they will dislike gift packages with significantly different prices. Furthermore, gift packages with prices similar to the preferred gift package, even if they are negative samples from users' exposure-based purchases, do not have their item labels set to 0; instead, a label smoothing operation can be performed to place them between 0 and 1. When a user makes a purchase on the same day, samples with large level deviations can be expanded into negative samples. In practical applications, a smoothed label is used to label the gift package purchased on that day, using the negative exponent of the level difference between the gift packages as the label. Suppose that purchased gift package P1 is level L1, and another undisplayed gift package P2 is level L2. In the original method, only P1 is a positive sample, and P2 is not included in the training samples. Now, gift package P2 will also be included in the training samples, labeled as exp(-abs(L1-L2)).

[0154] Optionally, after data filtering and augmentation, data cleaning can be performed on the sample data. For the gift package data of a period, for the positive sample gift packages of the same user, the negative sample labeled data is filtered out. That is, if a user buys gift package P1 on the first day but does not buy gift package P1 on the second day, the negative samples of not buying on the second day will not be used as training data.

[0155] Optionally, after obtaining the training samples after the above data filtering and augmentation, the model is trained based on the training samples to obtain the above refined model. The prediction goal of the refined model is to predict the user's purchasing behavior. Purchasing and not purchasing can be summarized as a binary classification problem, and the commonly used CTR prediction model can be used to train the prediction algorithm. Furthermore, a cross-entropy can be used as the objective function. Wide&Deep[4] and DeepFM[2] are used as prediction models.

[0156] Specifically, the training steps may include, but are not limited to: collecting user characteristics (users' spending power range), gift package item characteristics (resource consumption quantity), user historical interaction characteristics (users' spending behavior characteristics when purchasing gift packages), and contextual characteristics (characterizing the scenario information during user interaction, such as whether the interaction occurred on a weekday or weekend). Using user data from the activity period (e.g., one week or one month), 80% is selected as the training set, 20% as the validation set, and the actual effect is tested in the next activity period based on user clicks or purchases of the recommended gift packages. The input training data will be used to calculate the validation set loss according to the loss function mentioned above; the model with the minimum loss is saved for prediction.

[0157] Optionally, the above-mentioned method for recommending virtual props can be further applied to a personalized product recommendation system in a limited display scenario. This can be used, but is not limited to, for recommending discount packages, advertisements, and product props. A schematic diagram of the system workflow for the aforementioned personalized product recommendation system is shown below. Figure 5 As shown, the client's front-end display page is a product recommendation page visible to the user. When a user's front-end request 502 arrives, the recommendation page requests a back-end service 504. The back-end service 504, based on the front-end request 502, requests the personalized product recommendation system 506. The personalized product recommendation system 506 calculates the optimal recommendation result 508 for each request and sends it to the front-end for display.

[0158] Optionally, the aforementioned method for recommending virtual props can be further applied to a recommendation system application in a limited display scenario. For example... Figure 6As shown, after a 502 request is triggered on the front-end display page, a 602 request accesses the ranking service deployed in the Kubernetes container via the HTTP network request protocol. The ranking service constructs features for the current request, including user features, item features, user historical interaction features, and context features, and then requests the model service. The model service returns a model score (604) to the ranking service. The ranking service ranks the items according to the model score (604) and combines it with user segmentation, returning the results to the front-end, and recording the current features and model score (604), which are then transmitted to the feature-combining service. Simultaneously, the front-end records user feedback behavior, such as whether a click / purchase occurred after exposure, and also reports this to the feature-combining service. The feature-combining service records the recorded features and user feedback together and stores them in the file system for model training.

[0159] To further illustrate the above front-end display page, for example... Figure 3 As shown, virtual item 302 on the left is a fixed-position, fixed-price, limited-supply recommended item, while virtual item 308 on the right is a recommended gift pack containing a discount, available only on the day of purchase. It should be noted that the recommended item on the right has a limited display period. Here, we must consider not only user purchasing preferences but also the opportunity loss users might incur when making their choices. Even if the gift pack isn't the user's favorite, choosing not to purchase it would result in missing out on a discount.

[0160] It is understood that in the specific embodiments of this application, data such as user information are involved. When the above embodiments of this application are applied to specific products or technologies, user permission or consent is required, and the collection, use and processing of related data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

[0161] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0162] According to another aspect of the embodiments of this application, a virtual item recommendation apparatus for implementing the above-described virtual item recommendation method is also provided. For example... Figure 7 As shown, the device includes:

[0163] The first acquisition unit 702 is used to acquire the first historical consumption information and the second historical consumption data of the user account. The first historical consumption information is used to represent the consumption behavior triggered by the user account in the virtual game, and the second historical consumption information is used to represent the consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time period.

[0164] The second acquisition unit 704 is used to acquire the user account's price range preference in the virtual game based on the first historical consumption information, wherein the price range preference is used to represent the range of resources that the user account prefers to consume when triggering consumption behavior;

[0165] The third acquisition unit 706 is used to acquire the user account's virtual item preference in the virtual game based on the second historical consumption information, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behavior for any virtual item;

[0166] The fourth acquisition unit 708 is used to combine price range preference and virtual item preference to acquire the target virtual item to be pushed to the user's account. The target virtual item is a virtual item that meets the price range preference and virtual item preference among at least two virtual items.

[0167] As an optional solution, the third acquisition unit 706 includes:

[0168] The preference analysis module is used to perform preference analysis on the second historical consumption information to obtain the user account's preference degree for each virtual item in any virtual item. The preference degree is used to indicate the probability that the user account will purchase each virtual item.

[0169] The first determining module is used to determine virtual item preferences based on the preference level of each virtual item.

[0170] As an optional solution, the fourth acquisition unit 708 includes:

[0171] The second determining module is used to determine a first set of virtual items that meet the price range preference from at least two virtual items, and to determine the first virtual item with the highest preference from the first set of virtual items;

[0172] The third determining module is used to determine the first virtual item as the target virtual item; or,

[0173] The fourth determining module is used to determine a second set of virtual items from at least two virtual items whose preference degree is greater than or equal to a preset preference threshold, and to determine a second virtual item from the second set of virtual items that meets the price range and consumes the most resources.

[0174] The fifth determination module is used to determine the second virtual item as the target virtual item.

[0175] As an optional solution, the device also includes:

[0176] The display module is used to obtain the target virtual item to be pushed to the user's account after combining price range preference and virtual item preference, and then display the target item tag corresponding to the target virtual item. The target item tag includes a first item tag and a second item tag. The first item tag is used to represent the item attribute of the target virtual item, and the second item tag is used to represent the amount of resources consumed by the target virtual item.

[0177] As an optional solution, the device also includes:

[0178] The adjustment module is used to adjust the target item label corresponding to the target virtual item to the reference item label corresponding to the reference virtual item in response to the item refresh request after the item label corresponding to the target virtual item is displayed in the first display area, and to display the reference item label. Wherein, if the target virtual item is the first virtual item, the reference virtual item is the second virtual item, and if the target virtual item is the second virtual item, the reference virtual item is the first virtual item.

[0179] As an optional solution, the preference analysis module includes:

[0180] The extraction submodule is used to extract features from the second historical consumption information to obtain the user features of the user account in the recent time interval. The user features are used to indicate the user account's preference features for consumption behavior triggered by any virtual item in the recent time interval, and the item features corresponding to any virtual item.

[0181] The input submodule is used to input user features and the item features corresponding to each virtual prop into the preference prediction model, perform preference prediction processing, and obtain the preference degree of each virtual prop output by the preference prediction model. The preference prediction model is a neural network model obtained by pre-training the model using positive and negative samples.

[0182] As an optional solution, the device also includes:

[0183] The first acquisition module is used to acquire the set of historical virtual items that have been recommended to the user account before inputting user features and item features into the preference prediction model. The set of historical virtual items includes the set of third virtual items that the user account has purchased and the set of candidate virtual items that the user account has not purchased.

[0184] The second acquisition module is used to determine the candidate virtual item with the highest resource consumption from the third virtual item set before inputting user features and item features into the preference prediction model, and to determine at least one virtual item from the third virtual item set whose resource consumption is less than the resource consumption of the candidate virtual item by a degree less than a first preset deviation threshold, and to determine at least one virtual item whose deviation is less than the first preset deviation threshold as the first positive sample.

[0185] The third acquisition module is used to determine, before inputting user features and item features into the preference prediction model, at least one virtual item from the candidate virtual item set whose resource consumption quantity deviates from the resource consumption quantity of the candidate virtual item by more than a second preset deviation threshold, and to determine the at least one virtual item whose deviation degree is greater than the second preset deviation threshold as the first negative sample.

[0186] The training module is used to train the initial preference prediction model based on the first positive sample and the first negative sample before inputting user features and item features into the preference prediction model, so as to obtain the trained preference prediction model.

[0187] As an optional solution, the device also includes:

[0188] The fourth acquisition module is used to acquire a fifth set of virtual items that have not been recommended to the user's account before inputting user features and item features into the preference prediction model;

[0189] The fifth acquisition module is used to determine, before inputting user features and item features into the preference prediction model, multiple virtual props from the fifth set of virtual props whose resource consumption is less than the resource consumption of the candidate virtual props and to determine the multiple virtual props whose deviation is less than the first preset deviation threshold as the second positive sample.

[0190] The sixth acquisition module is used to identify, before inputting user features and item features into the preference prediction model, multiple virtual items from the fifth set of virtual items whose resource consumption is greater than the resource consumption of the candidate virtual items by a degree greater than a second preset deviation threshold, and to identify the multiple virtual items whose deviation is greater than the second preset deviation threshold as the second negative samples.

[0191] The training module includes:

[0192] The training submodule is used to train the initial preference prediction model by taking the first positive sample and the second positive sample as positive samples and the first negative sample and the second negative sample as negative samples, so as to obtain the trained preference prediction model.

[0193] As an optional solution, the second acquisition unit 704 includes:

[0194] The data analysis module is used to analyze the first historical consumption information to obtain the first price range of the amount of resources consumed corresponding to the consumption behavior triggered by the user account in the virtual game. The first price range is used to indicate the preferred amount of resources consumed from the lowest to the highest amount of resources consumed.

[0195] The sixth determining module is used to determine the first price range as the user account's price range preference in the virtual game when the deviation between the highest and lowest resource consumption is less than the third preset deviation threshold.

[0196] The seventh determining module is used to determine the second price range as the user account's price range preference in the virtual game when the deviation between the highest and lowest resource consumption is greater than or equal to the third preset deviation threshold. The second price range is a sub-range of the first price range and is used to indicate the preference from the middle resource consumption to the highest resource consumption, where the middle resource consumption is greater than the lowest resource consumption.

[0197] According to another aspect of the embodiments of this application, an electronic device for implementing the above-described method for recommending virtual items is also provided. This electronic device may, but is not limited to, [providing a specific function or capability]. Figure 1 The client 102 or server 112 shown in the figure, in this embodiment, is an electronic device as the client 102 as an example, and further as follows: Figure 8 As shown, the electronic device includes a memory 802 and a processor 804. The memory 802 stores a computer program, and the processor 804 is configured to execute the steps in any of the above method embodiments via the computer program.

[0198] Optionally, in this embodiment, the aforementioned electronic device may be located in at least one of a plurality of network devices in a computer network.

[0199] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0200] S1, obtain the first historical consumption information and the second historical consumption data of the user account, wherein the first historical consumption information is used to represent the consumption behavior triggered by the user account in the virtual game, and the second historical consumption information is used to represent the consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time period.

[0201] S2, based on the first historical consumption information, obtain the user account's price range preference in the virtual game, where the price range preference is used to represent the range of resources the user account prefers to consume when triggering consumption behavior;

[0202] S3, based on the second historical consumption information, obtain the user account's virtual item preference in the virtual game, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behavior for any virtual item;

[0203] S4. Combining price range preference and virtual item preference, obtain the target virtual item to be pushed to the user's account. The target virtual item is a virtual item that meets the price range preference and virtual item preference among at least two virtual items.

[0204] Alternatively, as those skilled in the art will understand, Figure 8 The structure shown is for illustrative purposes only. Figure 8 This does not limit the structure of the aforementioned electronic devices. For example, the electronic device may also include components that are more... Figure 8 The more or fewer components shown (such as network interfaces, etc.), or having the same Figure 8 The different configurations shown.

[0205] The memory 802 can be used to store software programs and modules, such as the program instructions / modules corresponding to the virtual item recommendation method and apparatus in this embodiment. The processor 804 executes various functional applications and data processing by running the software programs and modules stored in the memory 802, thereby implementing the aforementioned virtual item recommendation method. The memory 802 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 802 may further include memory remotely located relative to the processor 804, and these remote memories can be connected to electronic devices via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. Specifically, the memory 802 may be used, but is not limited to, to store information such as price range preferences and virtual item preferences. As an example, such as... Figure 8 As shown, the memory 802 may include, but is not limited to, the first acquisition unit 702, the second acquisition unit 704, the third acquisition unit 706, and the fourth acquisition unit 708 in the virtual item recommendation device. Furthermore, it may include, but is not limited to, other module units in the virtual item recommendation device, which will not be elaborated upon in this example.

[0206] Optionally, the transmission device 806 described above is used to receive or send data via a network. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 806 includes a Network Interface Controller (NIC), which can be connected to other network devices and routers via a network cable to communicate with the Internet or a local area network. In another example, the transmission device 806 is a radio frequency (RF) module, used for wireless communication with the Internet.

[0207] In addition, the aforementioned electronic device also includes: a display 808 for displaying information such as price range preferences and virtual item preferences; and a connection bus 810 for connecting the various module components in the aforementioned electronic device.

[0208] In other embodiments, the aforementioned client or server can be a node in a distributed system, wherein the distributed system can be a blockchain system, which is a distributed system formed by connecting multiple nodes through network communication. The nodes can form a peer-to-peer network, and any form of computing device, such as a server, client, or other electronic device, can become a node in the blockchain system by joining this peer-to-peer network.

[0209] According to one aspect of this application, a computer program product is provided, comprising a computer program / instructions containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs various functions provided in embodiments of this application.

[0210] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0211] It should be noted that the computer system of the electronic device is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0212] A computer system includes a Central Processing Unit (CPU), which performs various appropriate actions and processes based on programs stored in Read-Only Memory (ROM) or loaded from RAM. ROM also stores various programs and data required for system operation. The CPU, ROM, and RAM are interconnected via a bus. Input / output interfaces (I / O interfaces) are also connected to the bus.

[0213] The following components are connected to the input / output interface: input sections including keyboards, mice, etc.; output sections including cathode ray tubes (CRTs), liquid crystal displays (LCDs), and speakers; storage sections including hard drives; and communication sections including network interface cards such as LAN cards and modems. The communication section performs communication processing via a network such as the Internet. Drives are also connected to the input / output interface as needed. Removable media, such as disks, optical discs, magneto-optical discs, semiconductor memories, etc., are installed on the drive as needed so that computer programs read from them can be installed into the storage section as required.

[0214] Specifically, according to embodiments of this application, the processes described in the various method flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs various functions defined in the system of this application.

[0215] According to one aspect of this application, a computer-readable storage medium is provided, wherein a processor of a computer device reads computer instructions from the computer-readable storage medium, and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative implementations described above.

[0216] Optionally, in this embodiment, the computer-readable storage medium described above may be configured to store a computer program for performing the following steps:

[0217] S1, obtain the first historical consumption information and the second historical consumption data of the user account, wherein the first historical consumption information is used to represent the consumption behavior triggered by the user account in the virtual game, and the second historical consumption information is used to represent the consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time period.

[0218] S2, based on the first historical consumption information, obtain the user account's price range preference in the virtual game, where the price range preference is used to represent the range of resources the user account prefers to consume when triggering consumption behavior;

[0219] S3, based on the second historical consumption information, obtain the user account's virtual item preference in the virtual game, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behavior for any virtual item;

[0220] S4. Combining price range preference and virtual item preference, obtain the target virtual item to be pushed to the user's account. The target virtual item is a virtual item that meets the price range preference and virtual item preference among at least two virtual items.

[0221] Optionally, in this embodiment, those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware of an electronic device. The program can be stored in a computer-readable storage medium, which may include: flash drive, read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0222] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0223] If the integrated units in the above embodiments are implemented as software functional units and sold or used as independent products, they can be stored in the aforementioned computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause one or more computer devices (which may be personal computers, servers, or network devices, etc.) to execute all or part of the steps of the methods of the various embodiments of this application.

[0224] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0225] In the several embodiments provided in this application, it should be understood that the recorded client can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between units or modules may be electrical or other forms.

[0226] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0227] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0228] The above are merely preferred embodiments of this application. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for recommending virtual items, characterized in that, include: Obtain first historical consumption information and second historical consumption data of a user account, wherein the first historical consumption information is used to represent the consumption behavior triggered by the user account in the virtual game, and the second historical consumption information is used to represent the consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time period; Based on the first historical consumption information, the price range preference of the user account in the virtual game is obtained, wherein the price range preference is used to represent the range of resources preferred to be consumed when the user account triggers a consumption behavior; Based on the second historical consumption information, the user account's virtual item preference in the virtual game is obtained, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behavior for any virtual item; By combining the price range preference and the virtual item preference, a target virtual item is obtained and pushed to the user account, wherein the target virtual item is one of the at least two virtual items that meets the price range preference and the virtual item preference.

2. The method according to claim 1, characterized in that, The step of obtaining the user account's virtual item preferences in the virtual game based on the second historical consumption information includes: Preference analysis is performed on the second historical consumption information to obtain the user account's preference degree for each virtual item among the virtual items, wherein the preference degree is used to indicate the probability that the user account will purchase each virtual item; The virtual item preference is determined based on the preference level of each virtual item.

3. The method according to claim 2, characterized in that, The step of combining the price range preference and the virtual item preference to obtain the target virtual item to be pushed to the user account includes: From the at least two virtual items, determine a first set of virtual items that matches the price range preference, and from the first set of virtual items, determine the first virtual item with the highest preference. The first virtual item is identified as the target virtual item; or... From the at least two virtual items, determine a second set of virtual items whose preference degree is greater than or equal to a preset preference threshold, and from the second set of virtual items, determine the second virtual item that meets the price range and consumes the most resources; The second virtual item is identified as the target virtual item.

4. The method according to claim 1, characterized in that, After combining the price range preference and the virtual item preference to obtain the target virtual item to be pushed to the user account, the method further includes: Display the target item tag corresponding to the target virtual item, wherein the target item tag includes a first item tag and a second item tag, the first item tag is used to represent the item attribute of the target virtual item, and the second item tag is used to represent the amount of resources consumed by the target virtual item.

5. The method according to claim 4, characterized in that, After displaying the item tag corresponding to the target virtual item in the first display area, the method further includes: In response to an item refresh request, the target item tag corresponding to the target virtual item is adjusted to the reference item tag corresponding to the reference virtual item, and the reference item tag is displayed. Wherein, if the target virtual item is a first virtual item, the reference virtual item is a second virtual item, and if the target virtual item is a second virtual item, the reference virtual item is the first virtual item.

6. The method according to claim 2, characterized in that, The step of performing preference analysis on the second historical consumption information to obtain the user account's preference for each virtual item among the virtual items includes: Feature extraction is performed on the second historical consumption information to obtain the user characteristics of the user account in the recent time interval, wherein the user characteristics are used to indicate the user account's preference characteristics for consumption behavior triggered by any virtual item in the recent time interval, and the item characteristics corresponding to any virtual item; The user features and the item features corresponding to each virtual item are input into the preference prediction model for preference prediction processing to obtain the preference degree of each virtual item output by the preference prediction model. The preference prediction model is a neural network model obtained by pre-training the model using positive and negative samples.

7. The method according to claim 6, characterized in that, Before inputting the user features and the item features into the preference prediction model, the method further includes: Obtain the set of historical virtual items that have been recommended to the user account, wherein the set of historical virtual items includes the third set of virtual items that the user account has purchased and the set of candidate virtual items that the user account has not purchased; From the third set of virtual items, the candidate virtual item with the highest resource consumption is determined, and from the third set of virtual items, at least one virtual item whose resource consumption deviates from that of the candidate virtual item by less than a first preset deviation threshold is determined, and the at least one virtual item whose deviation is less than the first preset deviation threshold is determined as the first positive sample. From the candidate virtual item set, at least one virtual item whose resource consumption quantity deviates from the resource consumption quantity of the candidate virtual item by more than a second preset deviation threshold is identified, and the at least one virtual item whose deviation quantity deviates from the second preset deviation threshold is identified as the first negative sample. Based on the first positive sample and the first negative sample, the initial preference prediction model is trained to obtain the trained preference prediction model.

8. The method according to claim 7, characterized in that, Before inputting the user features and the item features into the preference prediction model, the method further includes: Obtain the fifth set of virtual items that were not recommended to the user's account; From the fifth set of virtual items, identify a number of virtual items whose resource consumption is less than the resource consumption of the candidate virtual items, and determine the number of virtual items whose deviation is less than the first preset deviation threshold as the second positive sample. From the fifth set of virtual items, identify multiple virtual items whose resource consumption deviates from the resource consumption of the candidate virtual items by a greater degree than the second preset deviation threshold, and define these multiple virtual items with a deviation greater than the second preset deviation threshold as the second negative samples. The step of training the initial preference prediction model based on the first positive sample and the first negative sample to obtain the trained preference prediction model includes: The first positive sample and the second positive sample are used as the positive samples, and the first negative sample and the second negative sample are used as the negative samples. The initial preference prediction model is trained to obtain the trained preference prediction model.

9. The method according to any one of claims 1 to 8, characterized in that, The step of obtaining the user account's price range preference in the virtual game based on the first historical consumption information includes: Data analysis is performed on the first historical consumption information to obtain a first price range of resource consumption corresponding to the consumption behavior triggered by the user account in the virtual game, wherein the first price range is used to indicate the preferred resource consumption from the lowest to the highest resource consumption. If the deviation between the highest and lowest resource consumption is less than a third preset deviation threshold, the first price range is determined as the user account's price range preference in the virtual game. If the deviation between the highest and lowest resource consumption is greater than or equal to the third preset deviation threshold, the second price range is determined as the user account's price range preference in the virtual game. The second price range is a sub-range of the first price range and is used to indicate the preference from the intermediate resource consumption to the highest resource consumption, wherein the intermediate resource consumption is greater than the lowest resource consumption.

10. A device for recommending virtual items, characterized in that, include: The first acquisition unit is used to acquire the first historical consumption information and the second historical consumption data of the user account, wherein the first historical consumption information is used to represent the consumption behavior triggered by the user account in the virtual game, and the second historical consumption information is used to represent the consumption behavior triggered by the user account for any one of at least two virtual items in the virtual game within a recent time period. The second acquisition unit is used to acquire the price range preference of the user account in the virtual game based on the first historical consumption information, wherein the price range preference is used to represent the range of the amount of resources preferred to be consumed when the user account triggers a consumption behavior; The third acquisition unit is used to acquire the user account's virtual item preference in the virtual game based on the second historical consumption information, wherein the virtual item preference is used to represent the user account's preference for triggering consumption behavior for any virtual item; The fourth acquisition unit is used to combine the price range preference and the virtual item preference to acquire the target virtual item to be pushed to the user account, wherein the target virtual item is one of the at least two virtual items that meets the price range preference and the virtual item preference.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, wherein the program is executed by an electronic device to perform the method according to any one of claims 1 to 9.

12. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the method described in any one of claims 1 to 9.

13. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to execute the method described in any one of claims 1 to 9 through the computer program.

Images