Content recommendation method, device, computer equipment and computer readable storage medium

Abstract

The present application discloses a content recommendation method, device, computer equipment and computer-readable storage medium, which relates to the field of data processing technology. In a target game, the hobbies of a player are accurately located according to the player's portrait data in other games, and the player is recommended content that matches the hobbies indicated by the portrait data. The recommendation accuracy is high, and waste of resources is avoided. The method includes: in response to the player's access request to the target game, obtaining the player's portrait data; determining the target game type of the target game, multiple designated game types of multiple designated games, and obtaining multiple game similarities between the target game type and the multiple designated game types; using multiple game similarities to calculate multiple sample class variance data corresponding to each designated game, and obtaining multiple mapping class variance data; determining game content matching the multiple mapping class variance data in the target game, and recommending the game content to the player based on the target game.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a content recommendation method, device, computer equipment and computer-readable storage medium. Background Art

[0002] With the rise of the game industry, various stand-alone games and online games have become an indispensable part of people's lives. In recent years, game content has become increasingly rich, and the gameplay has gradually diversified. In order to enable players to quickly locate what they are interested in in the huge game scene, many games will select information, data and other content that players may be interested in, and recommend these contents to players, so that players can quickly obtain the content of interest.

[0003] In related technologies, games usually obtain players' registration information, such as name, gender, age, address, etc., query content that is of interest to people of related age or gender, or obtain unions, communities, etc. near the address filled in by the player, and recommend the obtained content as content to be recommended to the player for browsing, so that the player can grasp this information in a timely manner.

[0004] In the process of implementing this application, the applicant found that the related technology has at least the following problems:

[0005] Content recommendations based on the information provided by players when registering are rather general. The content recommended to players may not be of interest to them, and the accuracy of content recommendations is not high, resulting in a waste of recommendation resources. Summary of the invention

[0006] In view of this, the present application provides a content recommendation method, apparatus, computer device and computer-readable storage medium, the main purpose of which is to solve the problem that the current content recommendation based on the information provided by the player when registering is relatively general, the content recommended to the player may not be of interest to the player, the accuracy of the content recommendation is not high, and the recommendation resources are wasted.

[0007] According to the first aspect of the present application, a content recommendation method is provided, the method comprising:

[0008] In response to a player's access request to a target game, obtaining the player's portrait data, the portrait data comprising a plurality of sample class variance data corresponding to each of a plurality of specified games, the plurality of sample class variance data being generated based on game data generated by the player in the corresponding specified game;

[0009] Determine a target game type of the target game and multiple designated game types of the multiple designated games, and obtain multiple game similarities between the target game type and the multiple designated game types;

[0010] Calculating a plurality of sample class variance data corresponding to each of the specified games using the plurality of game similarities to obtain a plurality of mapping class variance data;

[0011] Game content matching the plurality of mapping-like variance data is determined in the target game, and the game content is recommended to the player based on the target game.

[0012] Optionally, before obtaining the player's portrait data in response to the player's access request to the target game, the method further includes:

[0013] Determine a plurality of candidate games to be accessed, and extract the candidate games having the game records of the player from the plurality of candidate games as the plurality of designated games;

[0014] For each of the plurality of designated games, receiving game data of the player uploaded by the designated game;

[0015] According to the multiple functional modules, the game data is divided into multiple sample data groups;

[0016] Based on the class variance algorithm, class variance calculation is performed on each sample data group in the multiple sample data groups to obtain multiple sample class variance data;

[0017] Obtaining a game identifier of the designated game, and using the game identifier to mark the plurality of sample class variance data;

[0018] Calculate the corresponding multiple sample class variance data for the multiple specified games respectively, and obtain the multiple sample class variance data corresponding to each specified game;

[0019] The plurality of sample class variance data corresponding to each designated game is used as the portrait data of the player.

[0020] Optionally, after using the plurality of sample class variance data corresponding to each designated game as the portrait data of the player, the method further comprises:

[0021] Continuously counting the application time of the player's portrait data;

[0022] When the application time reaches a time threshold, candidate games having game records of the player are re-extracted from the multiple candidate games as the multiple designated games, and new portrait data of the player is generated based on the re-determined multiple designated games.

[0023] Optionally, the using the multiple game similarities to calculate the multiple sample class variance data corresponding to each specified game to obtain multiple mapping class variance data includes:

[0024] Respectively determine the functional module to which each sample class variance data in the plurality of sample class variance data corresponding to each designated game belongs, and divide the sample class variance data belonging to the same functional module into the same group to obtain a plurality of functional data groups;

[0025] For each of the function data groups, read a plurality of target sample class variance data included in the function data group, and perform class variance calculation on the function data group according to the correspondence between the plurality of game similarities and the plurality of target sample class variance data to obtain mapping class variance data of the function data group;

[0026] The mapping class variance data is calculated for each of the functional data groups respectively to obtain the multiple mapping class variance data corresponding to the multiple functional data groups.

[0027] Optionally, performing class variance calculation on the function data group according to the correspondence between the multiple game similarities and the multiple target sample class variance data to obtain the mapping class variance data of the function data group includes:

[0028] Respectively determine the designated game to which each target class variance data in the plurality of target sample class variance data belongs, mark each target sample class variance data with the game similarity corresponding to the designated game, and establish a corresponding relationship between the plurality of game similarities and the plurality of target sample class variance data;

[0029] Calculating the products of the plurality of target sample class variance data and the corresponding game similarities respectively to obtain a plurality of products;

[0030] Calculating the sum of the multiple products and counting the number of games of the multiple specified games;

[0031] The ratio of the sum to the number of games is calculated, and the ratio is used as mapping class variance data of the functional data set.

[0032] Optionally, determining game content matching the plurality of mapping-like variance data in the target game includes:

[0033] For each mapping class variance data among the plurality of mapping class variance data, determining a target function module to which the mapping class variance data belongs;

[0034] Acquire a preset grade standard of the target functional module, and query the target grade to which the mapping class variance data belongs in the preset grade standard;

[0035] Extracting the level content corresponding to the target level in the target function module;

[0036] The level content corresponding to each mapping class variance data in the corresponding functional module is determined respectively to obtain a plurality of level contents, and the plurality of level contents are used as the game content.

[0037] Optionally, determining game content matching the plurality of mapping-like variance data in the target game includes:

[0038] Determine the data attribute of each mapping class variance data in the plurality of mapping class variance data, and divide the mapping class variance data with the same data attribute into the same group to obtain at least two variance data groups;

[0039] For each of the at least two variance data groups, based on the mapping-type variance data included in the variance data group, determine a data group weight of the variance data group to obtain at least two data group weights of the at least two variance data groups;

[0040] A target data group weight is determined among the at least two data group weights, a target data attribute of the target data group weight is determined, and content marked with the target data attribute in the target game is used as the game content, wherein the target data group weight is greater than other data group weights among the at least two data group weights except the target data group weight.

[0041] Optionally, the method further comprises:

[0042] When the access request indicates registration in the target game, obtaining the player's portrait data, determining the game content according to the portrait data and making a recommendation; or,

[0043] When the access request indicates to log in the target game, it is queried whether the player has a historical login record in the target game, and based on whether the historical login record exists, the game content is recommended or a login operation for the target game is performed for the player.

[0044] Optionally, the recommending the game content or performing a login operation for the target game for the player according to whether the historical login record exists includes:

[0045] If the historical login record does not exist, obtaining the player's portrait data, determining the game content according to the portrait data and making a recommendation;

[0046] If the historical login record exists, a login operation for the target game is performed for the player.

[0047] According to a second aspect of the present application, a content recommendation device is provided, the device comprising:

[0048] an acquisition module, configured to acquire, in response to a player's access request to a target game, portrait data of the player, the portrait data comprising a plurality of sample class variance data corresponding to each of a plurality of specified games, the plurality of sample class variance data being generated according to game data generated by the player in the corresponding specified game;

[0049] A determination module, configured to determine a target game type of the target game and multiple designated game types of the multiple designated games, and obtain multiple game similarities between the target game type and the multiple designated game types;

[0050] A calculation module, configured to calculate a plurality of sample class variance data corresponding to each of the specified games using the plurality of game similarities to obtain a plurality of mapping class variance data;

[0051] A recommendation module is used to determine game content matching the plurality of mapping class variance data in the target game, and recommend the game content to the player based on the target game.

[0052] Optionally, the device further comprises:

[0053] An extraction module, configured to determine a plurality of candidate games to be accessed, and extract the candidate games having game records of the player from the plurality of candidate games as the plurality of designated games;

[0054] A receiving module, configured to receive, for each designated game among the plurality of designated games, game data of the player uploaded by the designated game;

[0055] A splitting module, used for splitting the game data into multiple sample data groups according to the multiple functional modules;

[0056] The calculation module is further used to perform class variance calculation on each sample data group in the multiple sample data groups based on a class variance algorithm to obtain multiple sample class variance data;

[0057] A marking module, used to obtain a game identifier of the specified game, and use the game identifier to mark the plurality of sample class variance data;

[0058] The calculation module is further used to calculate the corresponding multiple sample class variance data for the multiple specified games respectively, to obtain the multiple sample class variance data corresponding to each specified game;

[0059] The determination module is further configured to use the plurality of sample class variance data corresponding to each designated game as the portrait data of the player.

[0060] Optionally, the device further comprises:

[0061] A statistics module, used to continuously count the application time of the player's portrait data;

[0062] The extraction module is further configured to, when the application duration reaches a duration threshold, re-extract candidate games having game records of the player from the multiple candidate games as the multiple designated games, and generate new portrait data of the player based on the re-determined multiple designated games.

[0063] Optionally, the calculation module is used to respectively determine the functional module to which each sample class variance data belongs among the multiple sample class variance data corresponding to each specified game, and divide the sample class variance data belonging to the same functional module into the same group to obtain multiple functional data groups; for each functional data group in the functional data groups, read the multiple target sample class variance data included in the functional data group, and perform class variance calculation on the functional data group according to the correspondence between the multiple game similarities and the multiple target sample class variance data to obtain mapping class variance data of the functional data group; calculate the mapping class variance data for each functional data group respectively to obtain the multiple mapping class variance data corresponding to the multiple functional data groups.

[0064] Optionally, the calculation module is used to respectively determine the designated game to which each target class variance data in the multiple target sample class variance data belongs, mark each target sample class variance data with the game similarity corresponding to the designated game, and establish a corresponding relationship between the multiple game similarities and the multiple target sample class variance data; respectively calculate the products of the multiple target sample class variance data and the corresponding game similarities to obtain multiple products; calculate the sum of the multiple products, and count the number of games of the multiple designated games; calculate the ratio of the sum to the number of games, and use the ratio as the mapping class variance data of the functional data group.

[0065] Optionally, the recommendation module is used to determine, for each mapping class variance data in the multiple mapping class variance data, the target function module to which the mapping class variance data belongs; obtain a preset level standard for the target function module, and query the target level to which the mapping class variance data belongs in the preset level standard; extract the level content corresponding to the target level in the target function module; respectively determine the level content corresponding to each mapping class variance data in the function module to which it belongs, to obtain multiple level contents, and use the multiple level contents as the game content.

[0066] Optionally, the recommendation module is used to determine the data attributes of each mapping class variance data in the multiple mapping class variance data, divide the mapping class variance data with the same data attributes into the same group, and obtain at least two variance data groups; for each variance data group in the at least two variance data groups, determine the data group weight of the variance data group based on the mapping class variance data included in the variance data group, and obtain at least two data group weights of the at least two variance data groups; determine a target data group weight among the at least two data group weights, determine a target data attribute of the target data group weight, and use the content marked with the target data attribute in the target game as the game content, the target data group weight being greater than other data group weights among the at least two data group weights except the target data group weight.

[0067] Optionally, the acquisition module is further used to, when the access request indicates registration in the target game, acquire the portrait data of the player, determine the game content according to the portrait data and make a recommendation; or, when the access request indicates login in the target game, query whether the player has a historical login record in the target game, and based on whether the historical login record exists, recommend the game content or perform a login operation for the target game for the player.

[0068] Optionally, the acquisition module is used to obtain the player's portrait data if the historical login record does not exist, determine the game content according to the portrait data and make a recommendation; if the historical login record exists, perform a login operation for the target game for the player.

[0069] According to a third aspect of the present application, a computer device is provided, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the methods described in the first aspect when executing the computer program.

[0070] According to a fourth aspect of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of any one of the methods described in the first aspect are implemented.

[0071] By means of the above technical scheme, the present application provides a content recommendation method, apparatus, computer device and computer-readable storage medium. The present application generates player portrait data according to game data generated by the player in the game. When the player visits the target game, the portrait data is obtained, the target game type of the target game and the specified game types of multiple specified games visited by the player are determined, the game similarity between the target game type and the specified game type is obtained, the game similarity is used to calculate the multiple sample class variance data included in the portrait data to obtain multiple mapping class variance data, and the game content matching the multiple mapping class variance data is determined in the target game. Based on the target game, the game content is recommended to the player, so that in the target game, the player's hobbies can be more accurately located according to the player's portrait data in other games, and the content matching the hobbies indicated by the portrait data is recommended to the player. The accuracy of content recommendation is high, avoiding wasting recommendation resources.

[0072] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0074] Figure 1 A schematic diagram of a content recommendation method provided in an embodiment of the present application is shown;

[0075] Figure 2A A schematic diagram of a content recommendation method provided in an embodiment of the present application is shown;

[0076] Figure 2B A schematic diagram of a content recommendation method provided in an embodiment of the present application is shown;

[0077] Figure 2C A schematic diagram of a content recommendation method provided in an embodiment of the present application is shown;

[0078] Figure 3A A schematic diagram of the structure of a content recommendation device provided in an embodiment of the present application is shown;

[0079] Figure 3B A schematic diagram of the structure of a content recommendation device provided in an embodiment of the present application is shown;

[0080] Figure 3C A schematic diagram of the structure of a content recommendation device provided in an embodiment of the present application is shown;

[0081] Figure 4 A schematic diagram of the device structure of a computer device provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0082] The exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0083] The present application embodiment provides a content recommendation method, such as Figure 1 As shown, the method includes:

[0084] 101. In response to a player's access request to a target game, obtain the player's portrait data, where the portrait data includes a plurality of sample class variance data corresponding to each of a plurality of specified games, where the plurality of sample class variance data are generated based on game data generated by the player in the corresponding specified game.

[0085] 102. Determine a target game type of the target game and multiple designated game types of multiple designated games, and obtain multiple game similarities between the target game type and the multiple designated game types.

[0086] 103. Use multiple game similarities to calculate multiple sample class variance data corresponding to each specified game to obtain multiple mapping class variance data.

[0087] 104. Determine game content that matches the plurality of mapped variance data in a target game, and recommend the game content to a player based on the target game.

[0088] The method provided in the embodiment of the present application generates player portrait data based on game data generated by the player in the game, obtains the portrait data when the player visits a target game, determines the target game type of the target game and the specified game types of multiple specified games visited by the player, obtains the game similarity between the target game type and the specified game type, uses the game similarity to calculate multiple sample class variance data included in the portrait data to obtain multiple mapping class variance data, and determines game content matching the multiple mapping class variance data in the target game, and recommends the game content to the player based on the target game, so that in the target game, the player's hobbies can be more accurately located according to the player's portrait data in other games, and the content matching the hobbies indicated by the portrait data is recommended to the player. The accuracy of content recommendation is high, avoiding wasting recommendation resources.

[0089] The present application embodiment provides a content recommendation method, such as Figure 2A As shown, the method includes:

[0090] 201. Generate player portrait data.

[0091] The applicant realizes that the functions of the current content recommendation system in games are single and general, and usually use some user information or registration information provided by third-party platforms, such as the user's gender, age, address and some information filled in by the user on the platform, such as hobbies, etc. However, this information only roughly describes the basic situation of the user, and may not play a corresponding role when recommending relevant content to players based on this information in the game. The content recommended to players is not of interest to the players, and the accuracy of content recommendation is not high, resulting in a waste of recommendation resources.

[0092] Therefore, the present application proposes a content recommendation method, which generates player portrait data based on game data generated by the player in the game, obtains the portrait data when the player visits a target game, determines the target game type of the target game and the specified game types of multiple specified games visited by the player, obtains the game similarity between the target game type and the specified game type, uses the game similarity to calculate multiple sample class variance data included in the portrait data, obtains multiple mapping class variance data, and determines game content matching the multiple mapping class variance data in the target game, and recommends the game content to the player based on the target game, so that the player's hobbies can be more accurately located in the target game according to the player's portrait data in other games, and the content matching the hobbies indicated by the portrait data is recommended to the player. The accuracy of content recommendation is high, avoiding wasting recommendation resources.

[0093] The content recommendation method proposed in this application can be applied to a game platform, where multiple games are connected to the game platform, and these games are operated, mainly to solve the problems of cold start and user positioning. The game platform has its own user system, and can locate a user through the user's device ID (Identity Document, ID card identification number), user name and other information, and can calculate and store the data in the game registered, logged in or visited, and generate a series of universal special vectors as portrait data. In this way, when a user registers a new game, the user's portrait data is transmitted to the new game, so that the game can understand the user's habits and hobbies at the beginning and provide better services to the user. At the same time, in order to protect the user's portrait data, when the game platform provides portrait data to the new game, it will map the portrait data once, and convert the portrait data that the new game can use into a format that matches the new game, instead of sending all the portrait data directly to the new game, to avoid the leakage and waste of data resources.

[0094] In order to implement the technical solution in this application, it is necessary to generate player portrait data first. The portrait data includes multiple sample class variance data corresponding to each of multiple specified games, and the multiple sample class variance data are generated according to the game data generated by the player in the corresponding specified game. That is to say, the game platform will collect the game data generated by the player in the specified game, perform class variance calculation on the collected game data, and obtain multiple class variance values ​​as sample class variance data, so that the specified games registered, visited, and played by the player can correspond to multiple sample class variance data. The specific process of generating portrait data is as follows:

[0095] First, multiple candidate games to be connected are determined, and candidate games with game records of players are extracted from the multiple candidate games as multiple designated games. For each of the multiple designated games, the game data of the players uploaded by the designated game is received, and the game data is divided into multiple sample data groups according to multiple functional modules. Figure 2B, after the player logs in and enters the designated game A, B, C, D, game data will be generated in the designated game A, B, C, D, and the designated game A, B, C, D will upload these game data to the game platform, and the game platform will collect these game data for the subsequent portrait data generation process. Among them, the multiple functional modules can be multiple highly versatile main modules in the game, specifically including but not limited to the level module, the combat power module, the operation number module and the total recharge module, and the game data in each designated game can be grouped according to these modules. It should be noted that in addition to the main module described above, the multiple functional modules can also include some secondary modules, and the secondary modules can be divided into active modules and recharge modules; the game data divided into the active module can be divided according to the different functional modules of different games, such as the execution times of equipment enhancement, equipment upgrade, building upgrade, main task completion, branch task completion, etc. in the game can be divided into the active module; the game data divided into the recharge module can be the first recharge time, the number of recharges, etc. In the actual application process, a large number of functional modules can be defined, and according to the needs of the game and the computing power of the game platform, the portrait data can be calculated and sent. For the sake of versatility, about 4-8 functional modules with strong versatility can be selected. This application does not specifically limit the content and number of functional modules.

[0096] Subsequently, after obtaining multiple sample data groups, the game platform will perform class variance calculation on each of the multiple sample data groups based on the class variance algorithm to obtain multiple sample class variance data. Specifically, the class variance calculation can be performed based on the following formula 1 or formula 2:

[0097] Formula 1:

[0098] Formula 2:

[0099] In Formula 1 and Formula 2, σ x 2 and σ(x,y) are used to represent the value of the calculated sample class variance data, n is used to represent the number of game data included in the currently calculated sample data group, x i and i Used to represent the value of the game data included in the currently calculated sample data group, and It is used to represent the average value of the game data included in the sample data group currently calculated. It should be noted that, whether it is the standard variance or the covariance obtained by the class variance calculation, the purpose can be achieved. In other words, the sample class variance data records the difference between the user data and the standard average value during a calculation, which can be a relative value or a percentage, and the sample class variance data can be directly used for subsequent data mapping processing. It should be noted that, in general, the values ​​of the sample class variance data calculated by the sample data group related to the level (up to about 100) and the combat power (finally several trillion) in the game are very different, so in the actual application process, the obtained sample class variance data can also be processed, such as calculating the standard deviation of the sample class variance data, or normalizing the sample class variance data, etc., so as to reduce the gap between the sample class variance data.

[0100] After obtaining multiple sample class variance data through the above process, since these sample class variance data are generated based on the currently processed designated game, the game platform will obtain the game identifier of the designated game and use the game identifier to mark the multiple sample class variance data. Repeating the above process of generating sample class variance data, multiple sample class variance data corresponding to multiple designated games can be calculated respectively, and multiple sample class variance data corresponding to each designated game can be obtained, and the multiple sample class variance data corresponding to each designated game can be used as the player's portrait data.

[0101] It should be noted that after generating the player's portrait data, in order to maintain the accuracy of the recommended content, the game platform will continuously count the application time of the player's portrait data. When the application time reaches the time threshold, the candidate games with the player's game records are re-extracted from multiple candidate games as multiple designated games, and new player portrait data is generated based on the re-determined multiple designated games, so as to achieve periodic updates of the portrait data and maintain the universality of the portrait data.

[0102] 202. In response to a player's access request to a target game, obtain the player's portrait data.

[0103] In the embodiment of the present application, since the game platform generates portrait data for the player, these portrait data will be applied in the subsequent player's access to the game, that is, in response to the player's access request to the target game, the game platform will obtain the player's portrait data. Among them, the player's access request to the target game is actually divided into two types. One is a registration request. The registration request indicates that the player visits the target game for the first time and needs to recommend content based on the portrait data. Therefore, when the access request indicates to register in the target game, the game platform will obtain the player's portrait data, determine the game content according to the portrait data and recommend it. The other is a login request. Some games may only have a login page. Players also log in when they enter the game for the first time. When the game finds that the player has no login account, it will automatically register for the player. Therefore, when the access request indicates to log in in the target game, the game platform will query whether the player has a historical login record in the target game. According to whether there is a historical login record, the game content is recommended or the login operation of the target game is performed for the player. Specifically, if there is no historical login record, it means that the player is playing the game for the first time. Therefore, the game platform obtains the player's portrait data, determines the game content according to the portrait data and makes recommendations, so that the game platform can have a certain understanding of the player when the player registers. It should be noted that the player's portrait data does not have to be obtained at the same time as the player registers. The game platform can also obtain it after the novice guidance process is completed or when the target game actively applies to the game platform for content recommendation. If there is a historical login record, it means that the player is not logging in for the first time. In fact, the player's portrait data has been mapped to the target game before. Therefore, the game platform directly performs the login operation for the target game for the player.

[0104] 203. Determine a target game type of the target game and multiple designated game types of the multiple designated games, and obtain multiple game similarities between the target game type and the multiple designated game types.

[0105] In the embodiment of the present application, after obtaining the player's portrait data, since there is game similarity between games, the portrait data is generated based on other designated games registered by the player, and may not be fully compatible with the target game, and certain numerical adjustments are required. Therefore, the game platform will determine the target game type of the target game, multiple designated game types of multiple designated games, obtain multiple game similarities between the target game type and multiple designated game types, and subsequently adjust the portrait data according to the multiple game similarities, and map the portrait data to the target game so that the mapped value can match the actual game type and game content of the target game. Among them, the game similarity can be obtained from experience or calculated based on the game parameters in the game, such as calculating the proportion of the same functions and the same gameplay of the two games, and weighting these proportions according to the weights, and using the calculated value as the game similarity between the two games.

[0106] 204. Calculate multiple sample class variance data corresponding to each designated game using multiple game similarities to obtain multiple mapping class variance data.

[0107] In the embodiment of the present application, after obtaining multiple game similarities, the game platform will use multiple game similarities to calculate multiple sample class variance data corresponding to each specified game to obtain multiple mapping class variance data. The specific calculation process is as follows:

[0108] First, the game platform determines the functional module to which each sample class variance data in the multiple sample class variance data corresponding to each designated game belongs, and divides the sample class variance data belonging to the same functional module into the same group to obtain multiple functional data groups.

[0109] Subsequently, for each function data group in the function data group, the game platform reads the multiple target sample class variance data included in the function data group, and performs class variance calculation on the function data group according to the corresponding relationship between the multiple game similarities and the multiple target sample class variance data, and obtains the mapping class variance data of the function data group. Specifically, when performing class variance calculation, it is necessary to determine the specified game to which each target class variance data in the multiple target sample class variance data belongs, mark each target sample class variance data with the game similarity corresponding to the specified game, and establish the corresponding relationship between the multiple game similarities and the multiple target sample class variance data. Then, the product of the multiple target sample class variance data and the corresponding game similarity is calculated respectively to obtain multiple products, the sum of the multiple products is calculated, the number of games of the multiple specified games is counted, and the ratio of the sum to the number of games is calculated, and the ratio is used as the mapping class variance data of the function data group. In the actual application process, the mapping class variance data can be calculated based on the following formula 3:

[0110] Formula 3:

[0111] Among them, M represents the mapping variance data, n represents the number of games, a represents the game similarity, and X i Indicates the value of the target sample class variance data included in the currently calculated functional data group. It can be seen that the game similarity a is actually a weight, which is the percentage of influence of a certain game data on the target game. The value of a is a number between 0 and 1. When two games are completely similar, the value of a is 1. The initial value of the game similarity between each two games can be set to 1, and then the default game similarity can be updated based on experience or algorithms.

[0112] By re-executing the above process, the mapping class variance data can be calculated for each functional data group respectively, and a plurality of mapping class variance data corresponding to the plurality of functional data groups can be obtained.

[0113] 205. Determine, in the target game, game content that matches the plurality of mapping class variance data.

[0114] In the embodiment of the present application, after determining multiple mapping class variance data, the game platform will determine the game content that matches the multiple mapping class variance data in the target game, and then recommend these game contents to the players. Among them, there are two ways to determine the game content, as follows:

[0115] The first method is to directly obtain corresponding contents according to multiple mapping class variance data, and use all the obtained contents as game contents. That is, for each mapping class variance data in multiple mapping class variance data, determine the target function module to which the mapping class variance data belongs, obtain the preset level standard of the target function module, query the target level to which the mapping class variance data belongs in the preset level standard, extract the level content corresponding to the target level in the target function module, and determine the level content corresponding to each mapping class variance data in the function module to which it belongs, obtain multiple level contents, and use the multiple level contents as game contents. In fact, the above process is to evaluate the player's ability in multiple function modules according to multiple mapping class variance data, and obtain the content matching the player's ability as game content.

[0116] The second method is to comprehensively evaluate the player based on multiple mapping class variance data, determine which data the player tends to prefer, and recommend content according to the player's tendency. That is, determine the data attributes of each mapping class variance data in multiple mapping class variance data, divide the mapping class variance data with the same data attributes into the same group, and obtain at least two variance data groups. For each variance data group in at least two variance data groups, determine the data group weight of the variance data group based on the mapping class variance data included in the variance data group, obtain at least two data group weights of at least two variance data groups, determine the target data group weight in at least two data group weights, determine the target data attribute of the target data group weight, and use the content marked with the target data attribute in the target game as the game content. Among them, the target data group weight is greater than the other data group weights except the target data group weight in at least two data group weights, that is, the target data group weight is the largest data group weight.

[0117] The process described above is the process in which the game platform determines the game content by itself according to the player's portrait data. In the actual application process, the game platform can extract the label of the portrait data after obtaining the player's portrait data, that is, determine the function module that generates the portrait data, and select the target function label that can describe the value from the multiple function labels of the corresponding function module according to the value of the portrait data, and then obtain multiple target function labels that can describe the portrait data. Then, the game platform transmits these target function labels to the player based on the target game, so that the player can select the target function label that he likes from these target function labels, and the target game returns the selected target function label to the game platform. In this way, the game platform receives these selected target function labels, and when determining the game content that matches the multiple mapping class variance data, the data group weight of the variance data group indicated by the selected target function label can be increased, so that the variance data group is highlighted, so that the subsequent recommended game content includes the content that matches the selected target function label, and gives the player certain permissions in the process of recommending the game content, so that the recommended game content is more in line with the player's current interests and hobbies.

[0118] 206. Recommend game content to players based on the target game.

[0119] In the embodiment of the present application, after determining the game content, the game platform will recommend the game content to the player based on the target game. Among them, the game platform can directly send the game content to the target game, and the target game will recommend the content after the player enters, such as sending in-game emails, messages, etc. to the player; or the game platform can also directly push in-app messages to the player on the platform homepage, so that the player can also obtain the game content without logging into the target game. This application does not specifically limit the method of recommending game content to players.

[0120] In summary, the process of the content recommendation method proposed in this application is as follows:

[0121] like Figure 2C As shown, in response to the player's access request to the target game, the game platform determines whether the player enters the target game for the first time. If not, the login operation for the target game is directly performed; if yes, the game platform obtains the player's portrait data, maps the portrait data according to the game similarity, obtains multiple mapping class variance data, and extracts game content matching the multiple mapping class variance data in the target game, and recommends the game content to the player based on the target game.

[0122] The method provided in the embodiment of the present application generates player portrait data based on game data generated by the player in the game, obtains the portrait data when the player visits a target game, determines the target game type of the target game and the specified game types of multiple specified games visited by the player, obtains the game similarity between the target game type and the specified game type, uses the game similarity to calculate multiple sample class variance data included in the portrait data to obtain multiple mapping class variance data, and determines game content matching the multiple mapping class variance data in the target game, and recommends the game content to the player based on the target game, so that in the target game, the player's hobbies can be more accurately located according to the player's portrait data in other games, and the content matching the hobbies indicated by the portrait data is recommended to the player. The accuracy of content recommendation is high, avoiding wasting recommendation resources.

[0123] Further, as Figure 1 In a specific implementation of the method, the present application provides a content recommendation device, such as Figure 3A As shown, the device includes: an acquisition module 301, a determination module 302, a calculation module 303 and a recommendation module 304.

[0124] The acquisition module 301 is used to obtain the player's portrait data in response to the player's access request to the target game, wherein the portrait data includes a plurality of sample class variance data corresponding to each of a plurality of specified games, and the plurality of sample class variance data are generated according to the game data generated by the player in the corresponding specified game;

[0125] The determination module 302 is used to determine a target game type of the target game and multiple designated game types of the multiple designated games, and obtain multiple game similarities between the target game type and the multiple designated game types;

[0126] The calculation module 303 is used to calculate the multiple sample class variance data corresponding to each specified game using the multiple game similarities to obtain multiple mapping class variance data;

[0127] The recommendation module 304 is used to determine game content matching the plurality of mapping class variance data in the target game, and recommend the game content to the player based on the target game.

[0128] In specific application scenarios, such as Figure 3B As shown, the device also includes: an extraction module 305, a receiving module 306, a splitting module 307 and a marking module 308.

[0129] The extraction module 305 is used to determine a plurality of candidate games to be accessed, and extract the candidate games having the game records of the player from the plurality of candidate games as the plurality of designated games;

[0130] The receiving module 306 is used to receive the game data of the player uploaded by each of the multiple designated games;

[0131] The splitting module 307 is used to split the game data into multiple sample data groups according to the multiple functional modules;

[0132] The calculation module 303 is further used to perform class variance calculation on each sample data group in the multiple sample data groups based on a class variance algorithm to obtain multiple sample class variance data;

[0133] The marking module 308 is used to obtain the game identifier of the specified game, and use the game identifier to mark the plurality of sample class variance data;

[0134] The calculation module 303 is further used to calculate the corresponding multiple sample class variance data for the multiple specified games respectively, to obtain the multiple sample class variance data corresponding to each specified game;

[0135] The determination module 302 is further configured to use the plurality of sample class variance data corresponding to each designated game as the portrait data of the player.

[0136] In specific application scenarios, such as Figure 3C As shown, the device further includes: a statistical module 309.

[0137] The statistics module 309 is used to continuously count the application time of the player's portrait data;

[0138] The extraction module 305 is also used to re-extract the candidate games with the player's game records from the multiple candidate games as the multiple designated games when the application duration reaches a duration threshold, and generate new portrait data of the player based on the re-determined multiple designated games.

[0139] In a specific application scenario, the calculation module 303 is used to respectively determine the functional module to which each sample class variance data belongs among the multiple sample class variance data corresponding to each specified game, and divide the sample class variance data belonging to the same functional module into the same group to obtain multiple functional data groups; for each functional data group in the functional data groups, read the multiple target sample class variance data included in the functional data group, and perform class variance calculation on the functional data group according to the correspondence between the multiple game similarities and the multiple target sample class variance data to obtain mapping class variance data of the functional data group; calculate the mapping class variance data for each functional data group respectively to obtain the multiple mapping class variance data corresponding to the multiple functional data groups.

[0140] In a specific application scenario, the calculation module 303 is used to respectively determine the designated game to which each target class variance data in the multiple target sample class variance data belongs, mark each target sample class variance data with the game similarity corresponding to the designated game, and establish a corresponding relationship between the multiple game similarities and the multiple target sample class variance data; respectively calculate the product of the multiple target sample class variance data and the corresponding game similarities to obtain multiple products; calculate the sum of the multiple products, and count the number of games of the multiple designated games; calculate the ratio of the sum to the number of games, and use the ratio as the mapping class variance data of the functional data group.

[0141] In a specific application scenario, the recommendation module 304 is used to determine, for each mapping class variance data in the multiple mapping class variance data, the target function module to which the mapping class variance data belongs; obtain the preset level standard of the target function module, and query the target level to which the mapping class variance data belongs in the preset level standard; extract the level content corresponding to the target level in the target function module; respectively determine the level content corresponding to each mapping class variance data in the function module to which it belongs, obtain multiple level contents, and use the multiple level contents as the game content.

[0142] In a specific application scenario, the recommendation module 304 is used to determine the data attributes of each mapping class variance data in the multiple mapping class variance data, divide the mapping class variance data with the same data attributes into the same group, and obtain at least two variance data groups; for each variance data group in the at least two variance data groups, determine the data group weight of the variance data group based on the mapping class variance data included in the variance data group, and obtain at least two data group weights of the at least two variance data groups; determine the target data group weight in the at least two data group weights, determine the target data attribute of the target data group weight, and use the content marked with the target data attribute in the target game as the game content, and the target data group weight is greater than the other data group weights in the at least two data group weights except the target data group weight.

[0143] In a specific application scenario, the acquisition module 301 is also used to obtain the portrait data of the player when the access request indicates registration in the target game, determine the game content according to the portrait data and make recommendations; or, when the access request indicates login in the target game, query whether the player has a historical login record in the target game, and recommend the game content or perform a login operation for the target game for the player based on whether the historical login record exists.

[0144] In a specific application scenario, the acquisition module 301 is used to obtain the player's portrait data if the historical login record does not exist, determine the game content according to the portrait data and make recommendations; if the historical login record exists, perform the login operation for the target game for the player.

[0145] The device provided in the embodiment of the present application generates player portrait data based on game data generated by the player in the game, obtains the portrait data when the player visits a target game, determines the target game type of the target game and the specified game types of multiple specified games visited by the player, obtains the game similarity between the target game type and the specified game type, uses the game similarity to calculate multiple sample class variance data included in the portrait data to obtain multiple mapping class variance data, and determines game content matching the multiple mapping class variance data in the target game, and recommends the game content to the player based on the target game, so that in the target game, the player's hobbies can be more accurately located according to the player's portrait data in other games, and the content matching the hobbies indicated by the portrait data is recommended to the player. The accuracy of content recommendation is high, avoiding wasting recommendation resources.

[0146] It should be noted that for other corresponding descriptions of the functional units involved in the content recommendation device provided in the embodiment of the present application, reference can be made to Figure 1 and FIG. 2A to FIG. 2C The corresponding description in will not be repeated here.

[0147] In an exemplary embodiment, see Figure 4 , and also provides a computer device, which includes a bus, a processor, a memory and a communication interface, and may also include an input and output interface and a display device, wherein each functional unit can communicate with each other through the bus. The memory stores a computer program, and the processor is used to execute the program stored in the memory and execute the content recommendation method in the above embodiment.

[0148] A computer-readable storage medium stores a computer program, which implements the steps of the content recommendation method when executed by a processor.

[0149] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0150] Those skilled in the art will appreciate that the accompanying drawings are merely schematic diagrams of a preferred implementation scenario, and the modules or processes in the accompanying drawings are not necessarily required for implementing the present application.

[0151] Those skilled in the art will appreciate that the modules in the device in the implementation scenario can be distributed in the device in the implementation scenario according to the implementation scenario description, or can be changed accordingly and located in one or more devices different from the implementation scenario. The modules in the above implementation scenario can be combined into one module, or can be further split into multiple sub-modules.

[0152] The above application serial numbers are for description only and do not represent the advantages or disadvantages of the implementation scenarios.

[0153] The above disclosure only discloses several specific implementation scenarios of the present application. However, the present application is not limited thereto, and any changes that can be conceived by technicians in this field should fall within the scope of protection of the present application.

Claims

1. A content recommendation method, characterized in that: include: In response to a player's access request to a target game, obtaining the player's portrait data, the portrait data including sample class variance data corresponding to the player in a plurality of designated games, the plurality of sample class variance data being generated according to game data generated by the player in the corresponding designated games, wherein each of the designated games corresponds to a plurality of sample class variance data, the sample class variance data recording the difference between the user data of the player in the corresponding designated game and a standard average value when calculated; Determine a target game type of the target game and multiple designated game types of the multiple designated games, and obtain multiple game similarities between the target game type and the multiple designated game types, wherein the game similarities are obtained empirically or calculated based on game parameters in the game; The plurality of sample class variance data corresponding to each designated game are calculated by using the plurality of game similarities to obtain a plurality of mapping class variance data, wherein the plurality of sample class variance data corresponding to each designated game are calculated by using the plurality of game similarities to obtain a plurality of mapping class variance data, including: respectively determining the functional module to which each sample class variance data in the plurality of sample class variance data corresponding to each designated game belongs, dividing the sample class variance data belonging to the same functional module into the same group, and obtaining a plurality of functional data groups; for each functional data group in the functional data group, reading a plurality of target sample class variance data included in the functional data group, and performing class variance calculation on the functional data group according to the correspondence between the plurality of game similarities and the plurality of target sample class variance data, and obtaining the mapping class variance data of the functional data group; respectively calculating the mapping class variance data for each functional data group, and obtaining the plurality of mapping class variance data corresponding to the plurality of functional data groups; Game content matching the plurality of mapping-like variance data is determined in the target game, and the game content is recommended to the player based on the target game.

2. The method according to claim 1, characterized in that Before obtaining the player's portrait data in response to the player's access request to the target game, the method further includes: Determine a plurality of candidate games to be accessed, and extract the candidate games having the game records of the player from the plurality of candidate games as the plurality of designated games; For each of the plurality of designated games, receiving game data of the player uploaded by the designated game; According to a plurality of functional modules, the game data is divided into a plurality of sample data groups; Based on the class variance algorithm, class variance calculation is performed on each sample data group in the multiple sample data groups to obtain multiple sample class variance data; Obtaining a game identifier of the designated game, and using the game identifier to mark the plurality of sample class variance data; Calculate the corresponding multiple sample class variance data for the multiple specified games respectively, and obtain the multiple sample class variance data corresponding to each specified game; The plurality of sample class variance data corresponding to each designated game is used as the portrait data of the player.

3. The method according to claim 2, characterized in that After using the plurality of sample class variance data corresponding to each designated game as the player's portrait data, the method further includes: Continuously counting the application time of the player's portrait data; When the application time reaches a time threshold, candidate games having game records of the player are re-extracted from the multiple candidate games as the multiple designated games, and new portrait data of the player is generated based on the re-determined multiple designated games.

4. The method according to claim 1, characterized in that: The performing class variance calculation on the function data group according to the correspondence between the multiple game similarities and the multiple target sample class variance data to obtain the mapping class variance data of the function data group includes: Respectively determine the designated game to which each target class variance data in the plurality of target sample class variance data belongs, mark each target sample class variance data with the game similarity corresponding to the designated game, and establish a corresponding relationship between the plurality of game similarities and the plurality of target sample class variance data; Calculating the products of the plurality of target sample class variance data and the corresponding game similarities respectively to obtain a plurality of products; Calculating the sum of the multiple products and counting the number of games of the multiple specified games; The ratio of the sum to the number of games is calculated, and the ratio is used as mapping class variance data of the functional data set.

5. The method according to claim 1, characterized in that Determining the game content matching the plurality of mapping-type variance data in the target game comprises: For each mapping class variance data among the plurality of mapping class variance data, determining a target function module to which the mapping class variance data belongs; Acquire a preset grade standard of the target functional module, and query the target grade to which the mapping class variance data belongs in the preset grade standard; Extracting the level content corresponding to the target level in the target function module; The level content corresponding to each mapping class variance data in the corresponding functional module is determined respectively to obtain a plurality of level contents, and the plurality of level contents are used as the game content.

6. The method according to claim 1, characterized in that Determining the game content matching the plurality of mapping-type variance data in the target game comprises: Determine the data attribute of each mapping class variance data in the plurality of mapping class variance data, and divide the mapping class variance data with the same data attribute into the same group to obtain at least two variance data groups; For each of the at least two variance data groups, based on the mapping-type variance data included in the variance data group, determine a data group weight of the variance data group to obtain at least two data group weights of the at least two variance data groups; A target data group weight is determined among the at least two data group weights, a target data attribute of the target data group weight is determined, and content in the target game marked with the target data attribute is used as the game content, wherein the target data group weight is greater than other data group weights among the at least two data group weights except the target data group weight.

7. The method according to claim 1, characterized in that The method further comprises: When the access request indicates registration in the target game, obtaining the player's portrait data, determining the game content according to the portrait data and making a recommendation; or, When the access request indicates to log in the target game, it is queried whether the player has a historical login record in the target game, and based on whether the historical login record exists, the game content is recommended or a login operation for the target game is performed for the player.

8. The method according to claim 7, characterized in that The recommending the game content or performing a login operation for the target game for the player according to whether the historical login record exists includes: If the historical login record does not exist, obtaining the player's portrait data, determining the game content according to the portrait data and making a recommendation; If the historical login record exists, a login operation for the target game is performed for the player.

9. A content recommendation device, characterized in that: include: An acquisition module, configured to acquire the player's portrait data in response to a player's access request to a target game, wherein the portrait data includes sample class variance data corresponding to the player in a plurality of designated games, wherein the plurality of sample class variance data are generated based on the game data generated by the player in the corresponding designated games, wherein each of the designated games corresponds to a plurality of sample class variance data, and the sample class variance data records the difference between the user data of the player in the corresponding designated game and a standard average value when calculated; a determination module, configured to determine a target game type of the target game and multiple designated game types of the multiple designated games, and obtain multiple game similarities between the target game type and the multiple designated game types, wherein the game similarities are preset or calculated based on game parameters in the game; A calculation module, used for calculating the multiple sample class variance data corresponding to each designated game by using the multiple game similarities to obtain multiple mapping class variance data, wherein the calculation module is used for respectively determining the functional module to which each sample class variance data belongs in the multiple sample class variance data corresponding to each designated game, and dividing the sample class variance data belonging to the same functional module into the same group to obtain multiple functional data groups; for each functional data group in the functional data groups, reading the multiple target sample class variance data included in the functional data group, and performing class variance calculation on the functional data group according to the corresponding relationship between the multiple game similarities and the multiple target sample class variance data to obtain the mapping class variance data of the functional data group; respectively calculating the mapping class variance data for each functional data group to obtain the multiple mapping class variance data corresponding to the multiple functional data groups; A recommendation module is used to determine game content matching the plurality of mapping class variance data in the target game, and recommend the game content to the player based on the target game.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images